LTE L13 Radio Network Functionality

Transcription

1 LTE L13 Radio Network Functionality STUDENT BOOK LNA R13A LNA R13A

2 LTE L13 Radio Network Functionality DISCLAIMER This book is a training document and contains simplifications. Therefore, it must not be considered as a specification of the system. The contents of this document are subject to revision without notice due to ongoing progress in methodology, design and manufacturing. Ericsson shall have no liability for any error or damage of any kind resulting from the use of this document. This document is not intended to replace the technical documentation that was shipped with your system. Always refer to that technical documentation during operation and maintenance. Ericsson AB 2013 This document was produced by Ericsson. The book is to be used for training purposes only and it is strictly prohibited to copy, reproduce, disclose or distribute it in any manner without the express written consent from Ericsson. This Student Book, R13A supports course number LNA Ericsson AB 2013 LNA R13A

3 LTE Configuration Overview Table of Contents 1 LTE INTRODUCTION INTRODUCTION RADIO ACCESS NETWORK OVERVIEW LOGICAL ARCHITECTURE PRESENT FUNCTIONALITY IDLE MODE BEHAVIOR RADIO CONNECTION SUPERVISION POWER CONTROL, SCHEDULING AND LINK ADAPTATION CAPACITY MANAGEMENT AUTOMATED NEIGHBOR RELATION RADIO NETWORK CONFIGURATION LTE L13 CONFIGURATION OVERVIEW E CONFIGURATION TOOLS OVERVIEW RBS 6000 FAMILY OVERVIEW RBS TYPE VS. CONFIGURATION PROCESS MANAGED OBJECT MODEL RADIO NETWORK CONFIGURATION OVERVIEW RADIO NETWORK CONFIGURATION QOS CONFIGURATION SUMMARY IDLE MODE BEHAVIOR INTRODUCTION SYSTEM INFORMATION BROADCAST OF SYSTEM INFORMATION PLMN SELECTION AUTOMATIC MODE MANUAL MODE LNA R13A

4 LTE L13 Radio Network Functionality 3.3 ROAMING NETWORK CELL ACCESS RESTRICTION CELL BARRING AND CELL RESERVATION CELL SELECTION AND RESELECTION CELL SEARCH PROCEDURE CELL SELECTION PROCEDURE CELL RESELECTION PROCEDURE CELL RESELECTION EVALUATION PROCESS PRIORITY BASED CELL RESELECTION SPEED-DEPENDENT SCALING OF CELL RESELECTION TRACKING AREA UPDATE ACCESS CLASS RESTRICTIONS PAGING LOCATION REGISTRATION NORMAL REGISTRATION PERIODIC REGISTRATION PARAMETERS ATTRIBUTES OF MO EUTRANCELLFDD TO CONFIGURE IDLE MODE SUPPORT ATTRIBUTES OF MO EUTRANFREQRELATION TO CONFIGURE IDLE MODE SUPPORT PAGING RADIO CONNECTION SUPERVISION INTRODUCTION OVERVIEW OF RADIO CONNECTION SUPERVISION RADIO LINK MONITORING RADIO LINK MONITORING REQUIREMENTS PRINCIPLES OF RADIO CONNECTION SUPERVISION SUPERVISION OF UE IN RRC_CONNECTED STATE RRC CONNECTION RE-ESTABLISHMENT FEATURE DESCRIPTION Ericsson AB 2013 LNA R13A

5 LTE Configuration Overview 6 PARAMETERS POWER CONTROL, SCHEDULING AND LINK ADAPTATION INTRODUCTION QOS HANDLING SCHEDULING SCHEDULING DETAILS QOS AWARE SCHEDULER MINIMUM RATE PROPORTIONAL FAIR SCHEDULER UPLINK FREQUENCY SELECTIVE SCHEDULING UPLINK END-USER BITRATE SHAPING DELAY BASED SCHEDULING SEMI-PERSISTENT SCHEDULING TTI BUNDLING DL FREQUENCY SELECTIVE SCHEDULING RELATIVE PRIORITY SCHEDULING SERVICE SPECIFIC DRX CONNECTION SETUP RRM RELATED MEASUREMENTS IN LTE POWER CONTROL OPEN LOOP POWER CONTROL OPEN LOOP POWER CONTROL FOR RANDOM ACCESS DL POWER CONTROL UL POWER CONTROL LINK ADAPTATION CHANNEL PREDICTION PDSCH LINK ADAPTATION PDCCH LINK ADAPTATION PUSCH LINK ADAPTATION PUCCH LINK ADAPTATION POWER SPECTRAL DENSITY FOR POWER CONTROL OF PUCCH AND PUSCH LINK ADAPTATION OF INITIAL MESSAGES PHICH GROUPS LNA R13A

6 LTE L13 Radio Network Functionality 6.9 MAXIMUM TRANSMISSION POWER MIMO IN LTE DOWNLINK SCHEDULING AND LINK ADAPTATION LAYER 1 PROCESSING INTERFERENCE REJECTION COMBINING (IRC) PARAMETERS QOS PARAMETERS SCHEDULING PARAMETERS MINIMUM RATE PROPORTIONAL FAIR SCHEDULER PARAMETERS UL FREQUENCY SELECTIVE SCHDULING POWER CONTROL PARAMETERS PREAMBLE PARAMETERS MIMO PARAMETERS IRC PARAMETERS END USER BIT RATE SHAPING PARAMETERS DELAY BASED SCHEDULING PARAMETERS PARAMETERS FOR RELATIVE PRIORITY SCHEDULING DRX CONFIGURATION PARAMETERS PARAMETERS TTI BUNDLING CAPACITY MANAGEMENT CAPACITY MANAGEMENT INTRODUCTION ADMISSION CONTROL PRIVILEGED ACCESS DYNAMIC GBR ADMISSION CONTROL DESCRIPTION DIFFERENTIATED ADMISSION CONTROL DIFFERENTIATED ADMISSION CONTROL WITH UE AND BEARER PRE-EMPTION DYNAMIC QOS MODIFICATION OPERATOR DEFINED QCI FEATURE DESCRIPTION CONFIGURATION Ericsson AB 2013 LNA R13A

7 LTE Configuration Overview 8 CAPACITY LICENSES PARAMETERS PARAMETERS FOR ADMISSION CONTROL AFFECTED PARAMETERS FOR THE DYNAMIC GBR ADMISSION CONTROL FEATURE PARAMETERS INTRODUCED BY THE FEATURE OPERATOR DEFINED QCI CAPACITY LICENSES ARP BASED ADMISSION CONTROL DIFFERENTIATED ADMISSION CONTROL MOBILITY LTE MOBILITY LTE MEASUREMENTS INTRA LTE HANDOVER EVENT A INTER FREQUENCY MOBILITY IRAT MOBILITY COVERAGE TRIGGERED SESSION CONTINUITY COVERAGE TRIGGERED WCDMA IRAT HANDOVER SRVCC HANDOVER TO UTRAN SUBSCRIBER TRIGGERED MOBILITY DESCRIPTION IDLE MODE MOBILITY CONNECTED MODE MOBILITY CELL RESERVED FOR OPERATOR USE SERVICE TRIGGERED MOBILITY REDIRECT WITH SYSTEM INFORMATION UE LEVEL OSCILLATING HANDOVER MINIMIZATION CS FALLBACK WHY CS FALLBACK? EMERGENCY CALL FOR CS FALLBACK NEIGHBOR CELL RELATIONS LNA R13A

8 LTE L13 Radio Network Functionality 9 INTER-FREQUENCY LOAD BALANCING SUMMARY DESCRIPTION FEATURE BENEFITS: LOAD MEASUREMENT LOAD INFORMATION EXCHANGE CONFIGURATION PARAMETERS INTRA-LTE HANDOVER COVERAGE-TRIGGERED CDMA- EHRPD/WCDMA/GERAN/IF SESSION CONTINUITY SUBSCRIBER TRIGGERED MOBILITY SERVICE TRIGGERED MOBILITY REDIRECT WITH SYSTEM INFORMATION CS FALLBACK INTER FREQUENCY LOAD BALANCING SRVCC HANDOVER TO UTRAN AUTOMATED NEIGHBOR RELATIONS AUTOMATED NEIGHBOR RELATIONS ANR FOR LTE MEASUREMENTS INTRA-LTE/FREQUENCY ANR FUNCTION INTER-FREQUENCY & IRAT ANR FUNCTION ANR FOR UTRAN AND GERAN FUNCTION PERIODIC ANR MEASUREMENTS LIMIT ANR MEASUREMENT BENEFIT DESCRIPTION ENHANCEMENTS INDIRECTLY RELATED TO ANR PCI CONFLICT DETECTION PARAMETERS INTRODUCED ANR LTE PARAMETERS ANR UTRAN PARAMETERS Ericsson AB 2013 LNA R13A

9 LTE Configuration Overview 7.3 ANR GERAN PARAMETERS PCI CONFLICT REPORTING PARAMETERS ACRONYMS AND ABBREVIATIONS INDEX TABLE OF FIGURES LNA R13A

10 LTE L13 Radio Network Functionality Intentionally Blank Ericsson AB 2013 LNA R13A

11 LTE Configuration Overview 1 LTE Introduction Objectives After this chapter the participants will be able to: Explain the logical architecture of E-UTRAN and introduce Radio Functionality Detail the logical architecture of the Ericsson E-UTRAN List the Radio Functionality supported in the Ericsson E-UTRAN Figure 1-1: Objectives of Chapter 1 LNA R13A

12 LTE L13 Radio Network Functionality Intentionally Blank Ericsson AB 2013 LNA R13A

13 LTE Configuration Overview 1 Introduction This chapter gives an overview of the Ericsson LTE Radio Access Network, LTE RAN. It describes the architecture and gives an overview of the functionality. Evolved Packet System (EPS) in 3GPP Release 8 is based on a simplified network architecture compared to Release 6. The number of user-plane nodes is reduced from four in Release 6 (NodeB, RNC, SGSN and GGSN) to only two (e- NodeB and S-GW) in EPS. Only a Packet Switched (PS) domain is defined in LTE. This means that the traditionally Circuit Switched (CS) services will be carried by PS bearers. WCDMA GGSN CN SGSN RNC A flat architecture for optimized performance and cost efficiency MME LTE/SAE RNC SAE CN P/S-GW Moving RNC functions to e- NodeB NodeB NodeB e-nodeb e-nodeb UE UE Figure 1-2. Simplified Network Architecture. LNA R13A

14 LTE L13 Radio Network Functionality 2 Radio Access Network Overview 2.1 Logical Architecture The LTE Radio Access Network, called E-UTRAN consists of Radio Base Stations (RBS, in 3GPP called enodeb) Operations Support System for Radio and Core (OSS-RC) TEMS tools TEMS Cellplanner for LTE Core Network Other Management Systems Mun OSS-RC Network Management Environment S1 S1 Mul Mul Mun TEMS RBS X2 RBS S1 X2 RBS Radio Access Network Uu UE Uu RBS OSS-RC TEMS Radio Base Station Operation Support System Radio Core TEMS Optimization Solution Figure 1-3 Radio Network Overview, Logical Interfaces The core network, called Evolved Packet Core (EPC) is interconnected with E- UTRAN by means of the S1 interface. E-UTRAN and EPC together with the UEs is referred to as EPS. EPC manages EPS Bearer Services for user data between the UE and the Packet Data Network Gateway (PDN-GW). The RBS provides the radio resources. The key interfaces for user data are the S1 interface between enb and core network and the Uu between User Equipment (UE) and enb. Within the RAN, the enbs communicate with each other over the X2 and with EPC over S1. See Figure 1-3. Operation and Maintenance is handled through the embedded management in the RBS and the Operations Support System for Radio and Core (OSS-RC). TEMS provides the tools needed to plan, implement, and optimize 3G and 4G networks. Using tools from the TEMS portfolio during the transition from 2G and 3G to 4G will help save time and money. TEMS also provides all the tools necessary to maintain the current network Ericsson AB 2013 LNA R13A

15 LTE Configuration Overview To ease deployment and operations, Ericsson provides a set of tailored customerservice and site-solution products. 2.2 Present functionality The following Radio Functionality is included in the E-UTRAN and will be explained in the following chapters of this course EPS Bearer Service The term "EPS Radio Bearer Service" describes the overall connection between the UE and the Core Network edge node, PDN-GW. The EPS Bearer carries the end-to-end service and is associated with QoS (Quality of Service) attributes as decided by the operator. For user data, it maps down to a Radio Bearer from the UE to the enb, and an S1 transport bearer between the enb and the S-GW in the CN. Between the S-GW and the PDN-GW, a S5/S8 bearer is used to convey the transport between these nodes. The E-RAB is carried by a Radio Bearer between the UE and the RBS, and a user plane S1 Bearer. Figure 1-4 illustrates these relationships. E-UTRAN EPC Internet UE enb S-GW P-GW Peer Entity End-to-end Service EPS Bearer Service External Bearer E-RAB S5/S8 Bearer Data Radio Bearer S1 Bearer Radio S1 S5/S8 Gi Figure 1-4. EPS Bearer Concept User Plane. All services require a Signaling Connection to carry Radio Resource Control (RRC) signaling between the UE and enb and Non Access Stratum (NAS) signaling between the UE and MME. LNA R13A

16 LTE L13 Radio Network Functionality E-UTRAN EPC Internet UE enb MME P-GW Peer Entity NAS signaling AS signaling Signaling Radio Bearer S1 CP Radio S1 S5/S8 Gi Figure 1-5 Control Plane Concept. The NAS messages are carried between the UE and the enb using the Radio Resource Control (RRC) protocol on a Signaling Radio Bearer (SRB). They are transmitted between the enb and the MME using the S1 Application Protocol. The SRBs carrying RRC messages are carried by Logical Channels that are mapped onto a transport channel and scheduled together with the user data onto the physical resources (Radio Link) by the MAC layer, see Figure 1-6. UE RBS MME S/P-GW EPS Bearer Service (S1-UP) Radio Link Data Radio Bearer Traffic Channel RRC Signalling Channel S1 Signalling Bearer Signalling Radio Bearer NAS Signalling Connection (S1-CP) Transport Bearer (GTP) Figure 1-6 EPS Radio Access Bearer and NAS Signaling Connection Ericsson AB 2013 LNA R13A

17 LTE Configuration Overview 2.3 Idle mode behavior When the UE is not connected to the network its behavior must be still controlled. The UE will be able to roam between GSM, WCDMA and LTE as illustrated in CDMA2000 WCDMA IDLE MODE GSM LTE Figure 1-7 Idle mode behavior, PLMN & Cell Selection The operation of the UE in idle mode will have a serious impact on network performance and capacity. For example, the configuration of the paging procedure and the tracking areas affect the battery consumption and the uplink and downlink signaling load, cell (re-)selection performance affects the interference generated at connection setup etc. Idle mode behavior is managed by the system information that is sent on the Broadcast Control Channel (BCCH) in each cell. The system information contains parameters that control cell selection and reselection, paging, location registration, access and also parameters related to other functions. By using different parameter settings in system information, the operator is allowed to modify the service area and the cells that the UE can camp on. LNA R13A

18 LTE L13 Radio Network Functionality Typical Idle Mode tasks are listed in Figure 1-8. Monitor Paging Monitor System Information PLMN Reselection Cell Reselection Location Registration IDLE MODE LTE Figure 1-8 Idle Mode Behavior, Camp Normally Ericsson AB 2013 LNA R13A

19 LTE Configuration Overview 2.4 Radio Connection Supervision The Radio Connection Supervision (RCS) algorithm supervises the radio connection between the E-UTRAN and a UE in connected mode. The purpose of the supervision is to judge whether or not the E-UTRAN still has control over the UE. This supervision is only performed in the E-UTRAN for the uplink. Supervision of the downlink is carried out by a similar supervision function located in the UE. In the example in Figure 1-9, as the UE leaves the coverage of RBS1 the network must decide when to disconnect the resources. The setting of the parameters for this algorithm is a balance between dropped calls on the network and quality of service. WHY: Not to use resources for the UEs that experience bad radio condition Not to charge end user for a service that can not be used Avoid hanging resources in the network No Coverage Coverage Coverage UE loses radio contact! RBS 1 How: UE monitor RBS monitor RBS2 Figure 1-9 Radio Connection Supervision In the example the network could be configured to hold onto the resources for the UE until it reaches the coverage of cell 2. However there will be a break in transmission during the period of no coverage. LNA R13A

20 LTE L13 Radio Network Functionality 2.5 Power Control, Scheduling and Link Adaptation The key feature of the RBS is resource allocation and scheduling of the UEs, both in UL and in DL. In order to provide efficient resource usage from the RAN side and also to serve demands coming from different UEs scheduler takes in inputs from various sources. As Figure 1-10 illustrates, the inputs is coming from CN are QoS Class Identifier (QCI) and Allocation and Retention Priority (ARP). These are interpreted to local RAN QoS attributes via QoS Framework. That translation will result in the UE prioritization. QCIs ARPs QoS Parameters QoS Framework Scheduling Link Adaptation Power Control RBS Resource assignments UL/DL TPC commands Channel feedback Figure 1-10 Scheduling, QoS, LA, PC Overview Scheduler also works closely with Link Adaptation function which main task is to select a proper Transport Format (TF) based on SINR estimations, UE Power headroom and scheduled bandwidth. Power Control is used to minimize the transmitted power and to compensate for channel fading. It is also used to reduce intra cell interference and power consumption of the UE. Power Control bases its decisions on Channel predictions that are using UE measurements provided in Channel Feedback Reports (CFRs) Ericsson AB 2013 LNA R13A

21 LTE Configuration Overview 2.6 Capacity Management There are two algorithms responsible for capacity management in LTE: Admission Control Congestion Control License Control Admission control blocks new incoming calls as well as handover attempts when the load in the system is high, thus reducing the call-dropping probability. Admission control is used in both the uplink and downlink. The admission decision is based on air interface load, by using measurements of uplink interference, downlink output power as well as the actual number of users. The admission functionality is also capable of including priority, for instance emergency calls, in the admission decision. In the illustration in Figure 1-11 the last UE is blocked because the cell load has reached the defined admission limit. Admission limit Cell Load RBS Figure 1-11 Capacity Management (Admission Control) The setting of parameters associated with Admission control can have a serious effect on Network Capacity and revenue generated. Congestion control is used to resolve overload in both the uplink and the downlink. It uses Power and RSSI (Received Signal Strength Indicator) measurements. In case of overload, congestion control reduces bit rates of delay tolerant existing connections or as a second option, removes existing connections. LNA R13A

22 LTE L13 Radio Network Functionality In the example in Figure 1-12 the Cell load rises due to the increased power requirement of the UE that is moving away from the RBS. When this load reaches a defined limit the RBS must reduce it by delaying Best Effort (BE) packets. Eventually other users may need to be disconnected to reduce the load. Delay packets Cell Load Congestion limit RBS Figure 1-12 Capacity Management (Congestion Control) Again the setting of parameters associated with this functionality can have a serious impact of capacity and Quality of Service. E-UTRAN Mobility The following types of mobility are supported in the E-UTRAN: Inter LTE Handover (between MME pools) Intra LTE Handover (within an MME pool) o Intra enb Handover o Inter enb Handover X2 Handover S1 Handover Inter Frequency LTE Handover Coverage triggered Inter frequency Session Continuity Ericsson AB 2013 LNA R13A

23 LTE Configuration Overview Coverage triggered Inter Radio Access Technology (Inter- RAT) Session Continuity o GSM o WCDMA o CDMA 2000 X2 RBS 1 RBS 2 Figure E-UTRAN Mobility. LNA R13A

24 LTE L13 Radio Network Functionality 2.7 Automated Neighbor Relation Automated Neighbor Relation (ANR) in LTE is a licensed feature that automatically builds up and maintains a neighbor list used for handover. Automaticaly sets up neighbor relations when needed Supervise neighbor relations Remove unused neighbor relations Figure 1-14 Automated Neighbor Relation ANR adds neighbor relations to the cell neighbor list when User Equipment (UE) measurement reports indicate that a possible new neighbor relationship has been identified. When this occurs, the RBS requests the UE to report the unique Cell Global Identity (CGI) of the potential neighbor cell. Using this information, the RBS automatically creates a neighbor relation between the serving cell and the new neighbor cell and handover is facilitated. ANR also supervise the usage of the neighbor relation and it can remove neighbor relations that have not been used for a long period of time. The feature can be used together with manual optimization of neighbor lists, and is also able to remove neighbor cell relations which have not been used within a particular time period. In the current release ANR for Intra and Inter frequency LTE cells and ANR for UTRAN cells are supported Ericsson AB 2013 LNA R13A

25 LTE Configuration Overview Figure 1-15 CPI Store Figure 1-16 Radio Network Features Description in Alex LNA R13A

26

27 LTE Configuration Overview 2 Radio Network Configuration Objectives Configure the Radio Network in RBS6000 Explain the concept of cell and its relation to sector and antennae system in RBS6000 Recognize the Managed Objects related to radio network configuration Identify some basic parameters related to cell and cell relations Identify, and, if necessary, change QoS related parameters in RBS6000 Figure 2-1: Objectives LNA R13A

28 LTE L13 Radio Network Functionality 1 LTE L13 Configuration Overview In the figure below, the configuration (=integration) flow of an RBS 6000 is shown. Step 1 in the process has to be run at the RBS site. The remaining steps can either be run manually (on site or from OSS-RC) or with the Auto Integration feature. In the latter case the Base Station Integration Manager (BSIM) in OSS- RC is used. Configuration stages O&M Configuration + NW Synch Reference Sector + Antennae + Site/cabinet solutions S1 (and X2) Transport Cells (and neighbors) Make test session Configuration files/tools usual way Site Installation file Site Basic file Site Equipment file Define RBS6000 in OSS-RC ( OSS/ARNE ) Define S1 (and X2) ( OSS/BCM ) Add Radio Network Definitions ( OSS / BCM ) Unlock Cells RBS6000 Common OaM Infrastructure (COMINF) OSS-RC Server On-site engineer Figure 2-2: LTE Integration Flow As is obvious from the figure above, the objectives of the first two files (Site Installation and Site Basic files) are to connect the RBS to the OSS-RC for Operation and Maintenance access, and to provide a network synchronization reference. Site Equipment file configures the sectors, antennae system and the external alarms in the RBS. The rest of the configuration is performed from the OSS-RC, which include the definition of the S1- and the X2-interfaces, and cell configuration on the RBS. The RBS integration procedure and the tools used are described in detail in the configuration related chapter (Chapter 2). An overview of the tools is provided in Figure 1-5 below. The tools important for Configuration tasks are described briefly in the following subsections Ericsson AB 2013 LNA R13A

29 LTE Configuration Overview 1.1 E Configuration Tools Overview COMINF - DHCP - File Repository (SMRS) - O&M NTP server Etc... COMINF RBS On-site engineer - Element Manager (EMAS) NCLI COLI OSS-RC Server - ONE - ARNE -SMO - LTE Performance Recording - ASM - ALV - AMOS (MoShell) - Common Explorer BSIM BCM (integrated in BSIM) Etc... Figure 2-3 LTE RAN OaM Tools Element Manager (EM) Toolbox- or EMAS This is mainly a troubleshooting tool intended to be used on site. However, it can also be used remotely and it can be used for configuring a node, with the Graphical User Interface (GUI). NCLI Node Command Line Interface (NCLI) tool is a text-based tool that has access to the node database and the node alarm list. It is started via a telnet or a secure shell (ssh) session to the node. ARNE Add Remove Network Element (ARNE) is the OSS-RC tool for adding a node representation in the OSS-RC itself. BCM Bulk Configuration Management (BCM) can be divided into two parts; Bulk Configuration for Transport (BCT) and Bulk Configuration for Radio (BCR). BCM can be accessed either independently of BSIM, or an integrated interface in the BSIM (Base Station Integration Manager) tool described below. LNA R13A

30 LTE L13 Radio Network Functionality AMOS This is a command line interface that is commonly known as MoShell. The tool targets configuration management, fault management and performance management. BSIM BSIM is the preferred configuration tool for LTE. Since it has such a big role in the integration procedure, it is described in an own subsection below. How to use BSIM is described in detail in the RBS6000 Configuration chapter BSIM Base Station Integration Manager (BSIM) has two functions: Adding new erbs to OSS (including configuring for auto integration). Auto integrating nodes (erbs and RBS) to OSS when the node is taken into service. Auto integrate is a function which removes manual steps when taking a node into service. When a node (which is configured for auto integration) is MIB synchronized with OSS, OSS automatically performs steps to make the node ready for traffic. BSIM is a view in OSS Common Explorer. The BSIM view consists of the following pages: Add Node Use the Add Node page to add erbs nodes. Auto Integrate Use the Auto Integrate page to add RBSs, erbss that were not added to the OSS topology through BSIM or erbss that were added to the OSS topology through BSIM but without the Auto Integrate option selected. Auto Integrate is also used to view Not Connected nodes. BSIM is a license enabled feature Add New erbs to OSS The Add New erbs function adds up to 100 new erbss to OSS in one go. It performs the following main steps: Add the erbs to OSS topology Ericsson AB 2013 LNA R13A

31 LTE Configuration Overview Import transport and optionally radio configuration for the new erbs to a Planned Configuration. Optionally configure the erbs for auto integration. To use the Add New erbs function, set up a number of templates which BSIM will read. Templates are added to OSS using Utility Services. A BSIM template contains information, mostly MOs and attributes, that BSIM uses to add the new erbs. A sample of each kind of templates is delivered with BSIM. These templates can be used to make customized templates. The BSIM GUI displays a list of templates. The correct template is selected in the BSIM GUI to display substitution attributes for that template. A template is a complete script used for the application to be run in the OSS-RC (e.g, ARNE, BCM, etc). Most of the parameters (attributes) in the templates are common parameters for all tenodebs. Any parameter that is considered to be unique is specified in the template as a substitution variable/attribute. A substitution attribute is a value that is considered to be site (enodeb) specific, and thus must be specified by the BSIM user. Examples of substitution attributes are node name, IP address, and enodeb Id. When creating templates, put values that are shared for all nodes directly into the template and put node unique values in as substitution attributes. Substitution attributes are distinguished from other attributes by the percentage signs surrounding the attribute: %<ATTRIBUTE NAME>% Substitution attributes are displayed in the BSIM client and filled in for each node when the Add new erbs function is used Configuring for Autointegration Preparations for Auto Integrate can be selected as part of the Add new erbs function. The following options exist: Unlock Cells - unlock cells after the node is MIB synchronized, and the configuration is complete Create Configuration Version - create a node backup (CV) after the node is MIB synchronized Upgrade Package - a path to a software package which will be installed on the node LNA R13A

32 LTE L13 Radio Network Functionality Install licenses - installs licenses on the node which would control the capacity and the features available for that enodeb Site Basic - select a template used to generate a SiteBasic XML and stored on the SMRS server, used to configure the node Site Equipment - select a template used to generate a SiteEquipment XML and stored on the SMRS server, used to configure the node Site Installation - select a template used to generate a SiteInstall XML file to a user defined directory DHCP / DNS - select a template which will set up the DHCP / DNS servers with relevant information required during the auto integration Ericsson AB 2013 LNA R13A

33 LTE Configuration Overview 2 RBS 6000 FAMILY OVERVIEW enodeb functionality is implemented in Ericsson s solution with RBS6000 family base stations. However, RBS6000 is not only for LTE but also a solution for WCDMA nodeb and GSM base transceiver station (BTS) implementation. RBS 6000 family provides backwards-compatibility with the highly successful RBS 2000 (2G/BSS access) and RBS 3000 (3G/WCDMA access) product lines. It offers a seamless, integrated and environmentally friendly solution and a safe, smart and sound roadmap for whatever tomorrow holds. The RBS 6000 series is designed to support multiple radio technologies. All common GSM, WCDMA and LTE frequencies are supported in a single cabinet with common support equipment and can be mixed in virtually any combination. RBS 6102 RBS6101 RBS6201 RBS6202 RBS6601 RBS6301 RBS6302 mrbs prbs mrrus RRUS11 RRUS12 AIR Macro cabinets Main units Small cells Remote radios Figure 2-4: RBS 6000 Portfolio As can be seen, Ericsson has a product portfolio that is suited for all types of radio deployment. LNA R13A

34 LTE L13 Radio Network Functionality When a site is being decided on, there are various considerations taken into account. The number of sectors required, the baseband capacity requirement, the location of the site all become input on deciding how a site product would eventually look like. 1. Select radio 2. Select digital unit DUL or DUS RU RRU AIR 3. Select cabinet Figure 2-5: Full Freedom A brief introduction of the various products within the RBS6000 base station family is provided below. Note that the stated maximum configuration options are from the RBS product point of view, not just for LTE. The actual possibility for L13 (LTE) would depend on the type of RBS configuration (number of sectors, diversity characteristics and output power) desired and the cositing requirements, among other variables. For complete details, please check the CPI document: RBS Configurations RBS 6102 Large Outdoor RBS6102 is the outdoor macro RBS providing a complete radio site including transport equipment, site power and battery backup. The cabinet can house up to two radio shelves, thus designed for high capacity multistandard demands. For battery backup requirements there are three available options, i.e internal battery backup, the BBU (battery backup unit underneath the RBS) or the BBS (a separate high capacity battery backup system). Sectors: Maximum 12 sectors Capacity: 3x40 or 6 x 20 MHz LTE with 2xMIMO Nr of Radio Units: 12 RUs in two radio shelves Power feeding: VAC, -48VDC Ericsson AB 2013 LNA R13A

35 LTE Configuration Overview RBS 6101 Small Outdoor RBS 6101 is the compact outdoor macro RBS. The cabinet houses one radio shelf and has space for power, transport and a short battery backup. RBS 6101 can also be equipped as a high capacity Main Unit for main remote solutions. Sectors: Maximum 6 sectors Capacity: 3x20 MHz LTE with 2xMIMO Nr of Radio Units: 6 RUs Power feeding: VAC, -48VDC RBS 6201 Indoor RBS 6201 is an indoor cabinet that houses a macro configuration as well as the transport equipment and site power supply needed for an entire RBS site. The 6201 is a complete site, for high capacity multi-standard demands integrated into one cabinet. The cabinet can house up to two radio shelves that provide flexible radio capacity. For battery backup requirements, a battery backup unit underneath the RBS or a separate high capacity battery backup system is available. Sectors: Maximum 12 sectors Capacity: 3x40 or 6x20 MHz LTE with 2xMIMO Nr of Radio Units: 12 in two Radio shelves Power feeding: VAC, -48VDC, -60VDC, +24VDC RBS 6202 Indoor RBS 6202 is an indoor macro RBS in the RBS6000 family. It is a complete RBS within one radio subrack, which can either be in a cabinet or mounted on a 19- inch rack. Up to six RUs and two DUs maybe present in the RBS RBS 6601 Main Remote The RBS 6601 is a main-remote RBS and a member of the RBS 6000 family. The Main Remote concept provides the same high-performance network capabilities as Macro base stations, but with lower power consumption and less site requirements. The RBS 6601 consists of an indoor main unit and a number of RRUs designed to be located near the antenna. An optical fiber cable connects each RRU to the main unit. The maximum length of the optical fiber connecting the main unit and an RRU is 40 km. The normal LTE deployment of the RBS6601 would have up to three sectors. The number of RRUs supported depends on the available configurations. LNA R13A

36 LTE L13 Radio Network Functionality Sectors: Maximum 6 sectors Capacity: 3x40 MHz LTE with 2xMIMO Nr of Radio Units: Power feeding: -48VDC RBS 6301 Main Remote The RBS 6301 is also a main-remote RBS and a member of the RBS 6000 family. The RBS 6301 consists of an outdoor main unit and a number of Remote Radio Units (RRUs). An optical fiber cable and DC cable connect each RRU to the main unit. Optical fiber cables are available in standard lengths, from a few meters up to several hundred meters. 2.2 RBS Type vs. Configuration Process From a configuration point of view, no matter which RBS product is applicable, the procedure is the same, as illustrated in Figure 1-4 earlier. The parameter values for various attributes might be different, not only because they are of different RBS types, but also because of the configuration options being used, antenna related parameters, etc. Managed Object Model (MOM) used for all the different RBS types is the same, making the operation and configuration easy for the operators Ericsson AB 2013 LNA R13A

37 LTE Configuration Overview 3 Managed Object Model Configuring the RBS6000 is a matter of creating or editing entries in each node s database (called MIB, Managed Information Base). The MIB is described by a model called MOM (Managed Object Model). The MOM is built up of different classes, so-called Managed Object Classes (MOC), which represents some kind of resource in a node (e.g. hardware, software or configuration). For example, an operator wants to be able to configure one of the Ethernet ports of a node. In order to provide an interface to this port, a Managed Object Class called GigabitEthernet is defined. In that class there are parameters where the operators can, for example, define which VLANs that the Ethernet port should support. Since each MOC can be related to some resource in a node, the focus on configuration often comes down to which Managed Objects are connected to different functions and features. Configuration applications like NCLI, EM (the Containment View) and AMOS act directly on the local MIB in the node. OSS-RC configuration applications such as BSIM act on a remote instance of the MIB, with the possibility of running processes over the whole LTE RAN. An object oriented approach is used to model resources in RBS6000. The three layers needed to implement this model are described in the figure below. RBS6000 Node Resource Layer Management Adaptation Layer Configuration Service Managed Objects Management Information Base Figure 2-6 LTE RAN Node Resources Modeling LNA R13A

38 LTE L13 Radio Network Functionality Resource Layer The Resource Layer comprises physical and logical resources involved in carrying and processing the LTE traffic within an RBS6000 node. This layer is structured according to the hardware and software components within the node. The Resource Layer contains a number of internal, non operator-configurable resources. Consequently, only an abstraction of the Resource Layer is presented to the management applications Management Adaptation Layer The Management Adaptation Layer raises the abstraction level from the Resource Layer to a higher level suitable to be handled by management applications through the Service layer. Only resources that are needed for configuration and supervision are modelled. The Management Adaptation Layer makes possible the configuration of logical resources even if the underlying physical resource does not already exist on the node. This facilitates processes, and reduces resource downtime, during upgrades, reconfigurations and network extensions. The Management Adaptation Layer presents to the Service Layer a model of the node s resources based on Managed Objects. Each Managed Object is an instance of a Managed Object class defined in the Managed Object Model (MOM). The MOM is updated for every system release Service Layer (Configuration service) Management Information Base (MIB) In order to configure an RBS6000 node, management applications need to get access to the Managed Objects. The Service Layer implements a Management Information Base (MIB) that centrally stores on the node all Managed Object instances and the relationships between Managed Object instances, as well as providing access to these Managed Object instances Configuration Service The Service Layer provides a number of services for different purposes: Alarm Service, Product Inventory Service, Performance Monitoring Service and, of course, Configuration Service. The Configuration Service ensures that configuration applications such as EMAS, NCLI and OSS-RC configuration applications can configure MO instances in the MIB Ericsson AB 2013 LNA R13A

39 LTE Configuration Overview Access The Service Layer provides external access to the MIB and MOs via the Common Object Request Broker Architecture (CORBA) protocol carried over the Internetwork Inter-ORB Protocol (IIOP) protocol. Managed Object Model (MOM) Model showing the various classes available in each node. Managed Information Base (MIB) Subrack Subrack Slot Slot Personalised version of the MOM for a particular node in a network. Detailed description of all the managed objects with identifiers and attributes and relationships between them. Slot Slot = = 1 1 Subrack Subrack = = MS MS Slot Slot = = Figure 2-7 MOM vs. MIB Managed Object Model Building Managed Object Class An MO represents a resource in the node, either a physical resource such as a plug-in unit or a fan, or a logical resource such as a software program or a protocol. A number of MO classes exist to model all the resources needed for the node configuration and supervision. There are also parameters associated with the resource, which are called the MO attributes. The MO can be configured by setting suitable values for the MO attributes. The state and configuration of the resource represented by the MO can be monitored by reading these attribute values. There can also be MO actions related to the resource. For example, an MO representing an executable program might have an action called "restart". LNA R13A

40 LTE L13 Radio Network Functionality Relative Distinguished Name (RDN) <<MoClass>> GigaBitEthernet ID GigaBitEthernetId Actions setdscppbit() Attributes operationalstate availabilitystatus State Attributes Specific Attributes Performance Monitoring Attributes (Counters) actualspeedduplex dscppbitmap reservedby userlabel pmifinerrorslink1 pmifouterrorslink2 Common Attributes Figure 2-8 Managed Object (MO) MO Naming MOs are identified by means of a "naming attribute" or identifier. All MO classes must have a name-giving attribute called MoClassNameId, for example, SubrackId. This name is also referred to as the Relative Distinguished Name (RDN). The RDN is formed from the MO name and the value part of the namegiving attribute. Example of an RDN: SubrackId=1. A Local Distinguished Name (LDN) is a sequence of RDNs, which forms a unique name within the node. Example of an LDN: The MO Subrack has the name-giving attribute SubrackId. As in the previous example, the RDN of the MO could be SubrackId=1. The LDN of the same MO could be: ManagedElement=1,Equipment=1,Subrack=1. For system-created MOs, the value of this name-giving attribute is set automatically Ericsson AB 2013 LNA R13A

41 LTE Configuration Overview MO Attributes State Attributes Many MOs contain state information, which identifies whether the MO is available for use. This state information is defined by the following attributes: OperationalState The OperationalState attribute specifies whether the MO is OK (ENABLED) or failed (DISABLED). These values are set automatically and are propagated to the MO from the resource. The OperationalState is used by all MOs that can fail. The possible values of this attribute are: DISABLED ENABLED AdministrativeState The AdministrativeState attribute indicates whether the MO can be used. The possible values of this attribute are: LOCKED use of this MO is not allowed UNLOCKED use of this MO is allowed SHUTTING_DOWN no new users are accepted AvailabilityStatus The AvailabilityStatus attribute qualifies the OperationalState attribute. Only one value at the time can be assigned. The AvailabilityStatus is either an enumeration of the values specified below, or a bit-map (set valued integer) where at least the values defined below can occur. Other values not specified here can also occur. LNA R13A

42 LTE L13 Radio Network Functionality Common Attributes A number of common attributes exist, but they are not present in all MOs. Note: In this section the name-giving attribute is omitted. UserLabel The UserLabel attribute enables a user-friendly label (free text) to be placed on the MO instance. Note: This attribute cannot be used for identifying MOs since there is no requirement on this attribute to have unique values within the node or network. ReservedBy The ReservedBy attribute specifies the identity of the MO or MOs that have referenced this MO, and is used to keep track of relationships between MOs in the node. The referencing MO, which can be a ManagedObject type or a specific MO type, sets the ReservedBy attribute. Note: It is not possible to delete an MO that is reserved by another Interpretation of Attribute Properties Mandatory The attribute s initial value setting is mandatory and must be provided when creating an instance of the MO. Optional with initial value assigned The attribute s initial value setting is optional, that is, it can be provided when creating an instance of the MO. If it is not provided, the attribute is set automatically to its default value. The default value is specified after the data type. The only exception concerns MO references, that is, attributes of type moref, for which the default value "null" is assumed. Optional but with no initial value assigned The attribute s initial value is set automatically. ReadOnly The management tool can retrieve the attribute value by a get operation, but a set operation cannot change it. Note: A ReadOnly attribute cannot be set when creating an MO instance. It can only be set automatically Ericsson AB 2013 LNA R13A

43 LTE Configuration Overview ReadWrite The management tool can retrieve the attribute value by a get operation, and use a set operation to change it. Restricted The management tool can use a set operation to specify the attribute value when creating an MO instance, but cannot change it subsequently. It can also retrieve the attribute value by a get operation. Persistent The attribute value is stored persistently. NonPersistent The attribute value is not stored persistently. This applies primarily to state attributes such as operationalstate and availabilitystatus. Also, performance monitoring attributes are specified as nonpersistent. Notification The management tool provides information stating when the attribute value is changed automatically or by a set operation. NoNotification The management tool does not provide information about when the attribute value is changed automatically or by a set operation MO Actions Some MOs have defined actions that are used to execute particular operations related to the MO. Some actions require attributes to be provided as inputs. For example, the action assignipaddress that changes IP address on the MO EthernetLink requires that the user provides an ipaddress, a subnetmask and the broadcastaddress attributes during the execution. Some other actions do not involve any of the MO attributes. For example, the MO Licensing has an action setemergencystate that opens up licensed features and capacity restrictions. No attributes are asked for during the execution Relationships Between MOs Containment, Parent-Child LNA R13A

44 LTE L13 Radio Network Functionality In the MOM there are parent child relationships that define containment between MOs. This is illustrated in figure 2-5 below. <<MoClass>> <<MoClass>> ManagedElement ManagedElement <<MoClass>> <<MoClass>> ENodeBFunction ENodeBFunction <<MoContain>> <<MoContain>> 1..1 <<MoClass>> <<MoClass>> TransportNetwork TransportNetwork 0 24 <<MoClass>> <<MoClass>> EUtranCellFDD EUtranCellFDD Figure 2-9 Relationships between MOs: Containment, Parent-Child In this example the TransportNetwork MO is the child of the parent ManagedElement MO and the EUtranCellFDD MO is the child of the parent ENodeBFunction MO. It is not possible to delete an MO if it has children and it is not possible to create an MO if the parent does not exist. In terms of cardinality, there is one (1) and only one (1) parent for an MO, but an MO can have a number of children. The maximum number of children for an MO is defined in the MOM, and is related to resource or model limitations. Association In the MOM there can be associations between several MOs. Two types of associations can exist: Unidirectional association: object (b) has to be present in order to be able to create object (a). This is the most common type of association. In the figure below, we see that the IpAccessHostEt MO class has a unidirectional association with the IpInterface MO class. Indeed, an IpAccessHostEt MO instance cannot be created without a reference to an existing IpInterface MO instance. Bidirectional association: object (a) uses and/or invokes methods in object (b), but this object is not necessary for the creation of object (a). In the figure below, we see that the PlugInUnit MO class has a bidirectional association with the Program MO class. Indeed, a PlugInUnit MO instance can be created without a reference to an existing Program MO instance, and vice versa Ericsson AB 2013 LNA R13A

45 LTE Configuration Overview In the MOM these two kinds of relations can both be modeled as shown in figure below. <<MoClass>> <<MoClass>> PlugInUnit PlugInUnit 0..n <<MoAssociation>> 0..n <<MoClass>> <<MoClass>> Program Program <<MoClass>> <<MoClass>> IpAccessHostEt IpAccessHostEt <<MoAssociation>> <<MoClass>> <<MoClass>> IpInterface IpInterface Figure 2-10 Relationships between MOs: Association In term of cardinality, there is a need to specify the number of possible references from one MO class to another. An association which requires 1 and only 1 MO instance is indeed a unidirectional association, as shown in the figure above. Note: The cardinality of an MO is not the same as the maximum possible number of reserved MOs required to create the MO in the node. However, in most cases cardinality and maximum possible number of MOs are the same, and, if it is not the case, this is also mentioned in the description of the MO. For further information about the MOM structure, refer to the CPI documentation. LNA R13A

46 LTE L13 Radio Network Functionality 4 Radio Network Configuration Overview Before going into the configuration of the parameters and the configuration process, an overview of the radio network is made. The radio network configuration of the RBS6000 may be divided in the following steps, as shown in the figure. End RBS6000 OMC Radio network S1- and X2- Transport Radio and transport Define radio and transport MOs On-Site / OMC NPC Site Equipment Site Basic Basic Package Site Equipment RBS HW Site Basic Configuration O&M Access and NW Sync Basic CV A Basic SW package is loaded Start Figure 2-11 RBS Configuration Phases For the radio network configuration, the RUL/RUS/RRU, together with their antennae setup are taken into account. From a procedure point of view, the radio network configuration includes the Site Equipment configuration (with its corresponding file, when the sectors and their antennae are configured) and the Radio Network configuration (activated commonly through the OSS-RC/BSIM) Ericsson AB 2012 LZT R1A

47 Power Control, Scheduling and Link Adaptation 5 Radio network Configuration Figure below shows the Managed Object Model representations used for the Radio Network. From a configuration point of view, only a couple of the Managed Objects need to be created by the operator, others are created automatically. Only those that are mandatory to bring an enodeb into service are covered here. ManagedElement EnodeBFunction Rcs SecurityHandling EUtraNetwork Paging EUtranCellFDD or EUtranCellTDD SectorCarrier SectorEquipmentFunction EUtranFrequency ExternalENodeB Function UeMeasControl EUtranFreqRelation ExternalEUtranCellFDD or ExternalEUtranCellTDD EUtranCellRelation Created as a part of the ANR function Auto-created with default values Figure 2-12: Radio Network Managed Object Model..Cell setup As the figure indicates only the EUtranCellFdd/EUtranCellTdd and the SectorEquipmentFunction need to be created by the operator to bring a basic LTE network. The Automated Neighbor Relation (ANR) function will define the neighbor relations- (for which the DNS server and the allowed/disallowed enodebs in the enodebfunction MO are used). In the following subsections, attributes related to the following MOs are listed: EUTRAN cells - included in the EUtranCellFdd MO class. This MO needs to be created by the operator. Paging - included in the Paging MO class. This MO is autocreated with default values. Radio connection supervision - included in the Rcs MO class. This MO is auto-created with default values. Carrier details are included in the SectorCarrier MO class. LZT R1A Ericsson AB

48 LTE L13 Radio Network Functionality Sector functions - included in the SectorEquipmentFunction MO class. This MO is created as a part of the Site Equipment Configuration. Security handling - included in the SecurityHandling MO class. This MO is auto-created with default values. User Equipment (UE) measurement control - included in the UeMeasControl MO class. This MO is auto-created with default values. The generic MO operations to use are create, delete, and setattribute, using Element Manager or other user interfaces such as BCM or AMOS. For session continuity, it might be necessary to have frequency/cell relations to other systems. The diagram below shows how the eutrancellfdd or eutrancelltdd are related to the MOs representing other systems. The other systems could be Cdma2000, WCDMA UTRAN or GSM RAN (GERAN) access networks. The term Inter Radio Access Technology or IRAT is often used to describe this relation. UtraNetwork EUtranCellFdd or EUtranCellTdd GeraNetwork Utran Frequency Utran FreqRelation Geran FreqGroup Relation Geran FreqGroup Cdma2000Network Cdma2000 FreqBand External UtranCellFdd Utran CellRelation Cdma2000 FreqBand Relation GeranCell Relation Geran Frequency ExternalGeran Cell Cdma2000 Freq Cdma2000 FreqRelation ExternalCdma 2000Cell Cdma2000 CellRelation Figure 2-13 MOs related to IRAT interworking The details of the IRAT interworking are not described any further in this book Ericsson AB 2012 LZT R1A

49 Power Control, Scheduling and Link Adaptation EUtranCell Configuration Before the eutrancellfdd or the eutrancelltdd MO can be defined, the sector must already have been configured, which is done during the Site Equipment configuration. Note that most of the parameters listed below already have default values! The figure below shows the main areas of configuration in the EUtranCellFdd or EUtranCellTdd Managed Object. ManagedElement EnodeBFunction Rcs SecurityHandling EUtraNetwork Paging EUtranCellFDD or EUtranCellTDD SectorCarrier SectorEquipmentFunction Used to configure e.g. - Cell Identities - Bandwidth and Radio Channels EUtranCellFDD or EUtranCellTDD - Maximum RF Output Power - Cell User Capacity and QoS - Cell Availability - Scheduling and Interface Management - Cell Handover -... Figure 2-14: EUtranCellFDD or EUTRANCellTDD Cell Identities: The following attributes provide identifying information for the cell in the network: enodebplmnid The ENodeB Public Land Mobile Network (PLMN) ID that forms part of the ENodeB Global ID used to identify the node over the S1 interface. cellid RBS internal ID for the EUtranCell, must be unique in the RBS. Together with the RBS ID and PLMN this is a universally unique Cell ID. physicallayercellidgroup LZT R1A Ericsson AB

50 LTE L13 Radio Network Functionality Physical-layer cell IDs are grouped into 168 unique physical-layer cell ID groups, each group containing 3 unique subidentities. This attribute identifies the group. This attribute and the physicallayersubcellid are used to calculate physical layer cell ID (see 3GPP TS ) that is sent as part of the system information (see 3GPP TS ). physicallayersubcellid Physical-layer cell identities are grouped into 168 unique physical-layer cellidentity groups, each group containing 3 unique subidentities. This attribute identifies the subidentity within the group. This attribute and physicallyercellidgroup are used to calculate the physical layer cell identity (see 3GPP TS ) that is sent as part of the system information (see 3GPP TS ). sectorcarrierref Refers to the SectorCarrier MO which the EutranCell includes. Only one instance of the SectorCarrier MO class shall be configured per cell. tac Tracking area code for the EUTRAN cell. The TAC assists in paging the UE and is transmitted in System Information. When a UE registers itself in the network, the core network stores information about the tracking area where the registration is performed. This information is used, for example, to assist the UE paging. The tracking area update procedure is used by the UE to update the registration of its actual location in the network. The core network provides the UE with a list of tracking areas where the registration is valid. The UE performs a new registration, either after a certain time (periodic registration), or when it enters a new tracking area where the registration is no longer valid. The operator configures the TAC associated with each cell Bandwidth and Radio Channels The following attributes identify bandwidth and channel numbers for the cell: earfcndl: The mapping from channel number to physical frequency is for specified E- UTRA bands described in 3GPP specification TS The values that can be used depend on national, operator specific frequency allocation as well as from the supported frequency band(s) of the equipment in the enodeb. dlchannelbandwidth: Downlink channel bandwidth in the cell. Valid values include 1400 (1.4MHz), 3000 (3MHz), 5000 (5MHz), (10 MHz), (15 MHz) or (20 MHz), specifying the bandwidth in which the LTE is deployed Ericsson AB 2012 LZT R1A

51 Power Control, Scheduling and Link Adaptation earfcnul: Specifies the channel number for the central UL frequency. The mapping from channel number to physical frequency is for specified E-UTRA bands described in 3GPP TS The values that can be used depend on national, operator specific frequency allocation as well as from the supported frequency band(s) of the equipment in the enodeb. ulchannelbandwidth: Uplink channel bandwidth in the cell. Valid values include 1400, 3000, 5000, 10000, 15000, Note: Some bandwidth-related attribute values require proper licenses Maximum RF Output Power The maximum possible cell transmission power depends on a number of factors, including RU hardware capability, licensed power provisioning and the configuration parameters. pmaxservingcell: Pmax to be used in the cell. If not found, the UE applies the maximum power according to the UE capability. pzeronominalpucch: Nominal component of the UE transmit power for Physical Uplink Control Channel (PUCCH). pzeronominalpusch: Nominal component of the UE transmit power for Physical Uplink Shared Channel (PUSCH) Cell User Capacity and QoS The following attributes provide information the cell user capacity and quality of service: noofpucchcqiusers Number of Channel Quality Indicator (CQI) resources available on the Physical Uplink Control Channel (PUCCH) channel. noofpucchsrusers LZT R1A Ericsson AB

52 LTE L13 Radio Network Functionality Number of scheduling request resources available on the PUCCH channel. qcitableref Refers to an existing MO QciTable and assigns a QoS Class Identifier (QCI) table to a cell. The values of both attributes noofpucchsrusers and noofpucchcqiusers must be set according to the following rules: The values may not be set to 0. The maximum values depend on cell bandwidth. If the bandwidth is 1.4 MHz the max values are 207 SR and 216 CQI. If the bandwidth is > 1.4MHz the maximum values are 230 SR and 240 CQI. The values must not violate the limits on maximum number of PUCCH PRB per cell as defined in the CPI Control Channel Dimensioning Cell Availability The following attributes provide information about cell availability: acbarringinfo Contains all access barring information. administrativestate The administrative status of the cell. availabilitystatus The availability status of the cell. It contains details about operationalstate. This attribute is set by the software in the RBS. cellbarred Specifies if the cell is required for a specific purpose and should not be accessible by random user equipment. primaryplmnreserved Indicates if the primary PLMN ID in the cell is reserved for operator use. The primary PLMN ID is reserved if this attribute is set to true. Note: The parameter enodebplmnid in the parent ENodeBFunction MO will hold the value of the primary PLMN ID operationalstate Ericsson AB 2012 LZT R1A

53 Power Control, Scheduling and Link Adaptation The operational state. This attribute is set by the software in the RBS Scheduling and Interference Management The following attribute provides information about scheduling and interference management for the cell: ulinterferencemanagementactive Enables or disables uplink interfe2rence management. ulfrequencyallocationproportion Frequency resources that is allocated in UL expressed as a percentage of the configured bandwidth Frame Management The following attribute provides information about frame management for the cell: framestartoffset In the current release the value must be set to Additional UE Output Power Restrictions Additional Maximum Power Reduction (A-MPR) can be signalled by a Network Signalling value to the UE to meet additional requirements in specific deployment scenarios and specific frequency bands (bands 1, 2, 4, 10, 13, 35 and 36). The UE is then allowed to reduce the output power to fulfill special spectrum emission requirements. This power reduction will in some cases have severe impact on the PUCCH performance. To overcome this problem it is possible to shift the PUCCH region close to the center of the system bandwidth where the required power reduction is less severe. The following attributes provides information about the additional UE output power restrictions: networksignallingvalue Specifies the Network Signalling value to be broadcast in the cell. Default is NS_01 meaning that no additional back-off is required by the UE. Other supported values are: NS_03 indicating bands 2, 4, 10, 35 and 36, NS_04, NS_05 indicating band 1, and NS_07 indicating band 13. pucchoverdimensioning LZT R1A Ericsson AB

54 LTE L13 Radio Network Functionality Specifies the number of resource blocks that the PUCCH shall be shifted at each band edge. Supported values are 0 and 13. Default value is 0, which corresponds to no PUCCH overdimensioning. This attribute is intended to provide PUCCH overdimensioning for Band 13. If networksignallingvalue is set to NS_07 (indicating band 13), then the value of pucchoverdimensioning shall be set to 13. No PUCCH overdimensioning should be used for any other Frequency Band. For all other values of networksignallingvalue, the attribute pucchoverdimensioning should have the value Cell Handover The following attributes configure handover in the cell: minbestcellhoattempts Number of attempts for handover to a cell better than the serving one, before handover is attempted to the next best cell. If there is no next best cell in the UE report, handover to the best cell is attempted repeatedly. reservedby Contains a list of EutranCellRelation MOs that reserve this managed object Cell Reselection The following attributes configure cell reselection: systeminformationblock3 Contains cell reselection information common for intra-frequency, interfrequency, and/or inter-rat cell reselection. systeminformationblock6 Contains cell reselection information common for cell reselection towards UTRAN. systeminformationblock7 Contains cell reselection information common for cell reselection towards GERAN. systeminformationblock8 Contains cell reselection information common for cell reselection towards CDMA Ericsson AB 2012 LZT R1A

55 Power Control, Scheduling and Link Adaptation DRX for Connected UE Whether Discontinuous Reception (DRX) for connected UE is available or not can be checked/set with the attribute: drxactive Specifies/allows if the DRX is possible or not in the cell. Note that DRX is a licensed feature that is activated per cell. It should be activated with a valid license. The MOs Licensing and Drx should be consulted. If using the Service Specific DRX feature, the licence for ServiceSpecifilDRX MO class is also necessary Spatial Division Multiplexing Support Whether Spatial Division Multiplexing (SDM) is supported in the cell or not is found by checking the attribute: sdmactive Specifies if SDM for combined cell configuration is activated or not. If SDM is activated, it is possible to multiplex multiple UEs in different sector carriers in the same time and same frequency resource. Note: The SDM functionality for combined cell is not supported. If the parameter is set to true, it will not take effect Random Access This section covers configuration of random access functionality. rachrootsequence The first root sequence number for Random Access Channel (RACH) preamble generation. RACH root sequence is broadcast as a part of system information distribution and used for preamble detection. The definition for logical root sequence number can be found in 3GPP TS cellrange Defines the maximum distance from the RBS where a connection to a UE can be setup or maintained, or both. cfraenable This parameter is used to enable or disable Contention Free Random Access (CFRA). LZT R1A Ericsson AB

56 LTE L13 Radio Network Functionality Paging configuration The values of the attributes of the Paging MO class determine the paging characteristics. This MO is auto-created with default values. ManagedElement EnodeBFunction Rcs SecurityHandling EUtraNetwork Paging EUtranCellFDD or EUtranCellTDD SectorCarrier SectorEquipmentFunction Paging Used to configure e.g. - Default paging cycle - Delay and bandwidth used by paging functionality This MO is auto-created, with default values, by the system Figure 2-15: Paging defaultpagingcycle The attribute determines the Default Paging Cycle used by the RBS and broadcast in System Information (SIB2). This value is overridden by the UE specific paging cycle, if it is provided in incoming Paging from MME, and if it is smaller than the defaultpagingcycle. Setting a smaller value for the attribute defaultpagingcycle results in a shorter paging time, but a more "bursty" Paging traffic, since more UEs will be grouped to use the same PF. A larger value for the attribute results in longer paging times, but distributes the Paging traffic more evenly in time. maxnoofpagingrecords The number of UEs paged in a PO can not exceed the number specified by the value of the maxnoofpagingrecords attribute. If the received number of Paging messages from MME, targeting the same PO, exceeds maxnoofpagingrecords, the excess paging records are sent in the next PO in the Paging Cycle. Hence setting this parameter to a smaller value may cause extra delays for some Paging messages, but instead it limits the bandwidth used by paging. A larger value may require more of the RBS bandwidth but decreases the probability of extra delays for UEs. nb Ericsson AB 2012 LZT R1A

57 Power Control, Scheduling and Link Adaptation The value of the attribute determines the number of Paging Occasions (POs) in each Paging Frame (PF). Together with the attribute defaultpagingcycle it determines the paging time and the distribution of the Paging traffic in time. Setting the value of the nb attribute so that there are fewer POs in each PF will decrease the total number of available POs and hence the Paging traffic will be more "bursty". pagingdiscardtimer The attribute determines the maximum period of time a received Paging may be retained or queued in the RBS before it is discarded. The value of the pagingdiscardtimer should be the same as that set in the paging resend timer in the MME. Setting the attribute value too high will result in the RBS paging queue containing several instances of the same paging sent by the MME. Setting the value of the pagingdiscardtimer attribute too low will remove some paging messages from the RBS paging queue prematurely. This will result in unnecessary paging resend from the MME. Further information can be found in CPI document TA Planning Guideline. LZT R1A Ericsson AB

58 LTE L13 Radio Network Functionality RCS Configuration The attributes of the Rcs MO class determine the Radio Connection Supervision (RCS) characteristics. This MO is auto-created with default values. ManagedElement EnodeBFunction Rcs SecurityHandling EUtraNetwork Paging EUtranCellFDD or EUtranCellTDD SectorCarrier SectorEquipmentFunction Rcs Radio Connection Supervision The time a UE can be inactive before it is released This MO is auto-created, with default values, by the system Figure 2-16: Radio Connection Supervision tinactivitytimer The attribute determines the time that a UE can be inactive before it is released. Setting a suitable value for the attribute tinactivitytimer improves performance. The default value is 61 seconds. Setting the attribute value too low results in too frequent setups and releases of UEs. Setting the value too high also results in some UEs using more system resources than needed Ericsson AB 2012 LZT R1A

59 Power Control, Scheduling and Link Adaptation Sector configuration SectorCarrier represents the usage of the resources that the MO SectorEquipmentFunction refers to. Examples include power and antenna. ManagedElement EnodeBFunction Rcs SecurityHandling EUtraNetwork Paging EUtranCellFDD or EUtranCellTDD SectorCarrier SectorEquipmentFunction SectorCarrier Used to configure e.g. - NoOfRxAntennas - NoOfTxAntennas - partofsectorpower Figure 2-17: Sector Carrier maximumtransmissionpower Maximum possible power at the antenna reference point, for all downlink channels in all TX branches used simultaneously in the SectorCarrier. -1 is used as an undefined value. ReadOnly. noofrxantennas The number of antennas that can be used for receiver diversity nooftxantennas The number of antennas that can be used for MIMO or for transmitter diversity Note: Some antenna configurations require licenses. partofsectorpower (in percentage) LZT R1A Ericsson AB

60 LTE L13 Radio Network Functionality The requested part of the total power in the SectorEquipment that should be allocated for the sectorcarrier. The output power is evenly distributed over the antenna connectors used for TX transmission that are allocated for the SectorCarrier. If the total amount of power per antenna connector of a SectorEquipment is over 100% when the cell using the SectorCarrier is unlocked, an alarm will be generated. prsenabled If true, the Positioning Reference Signal will be sent from this sectorcarrier. As obvious from the MOM diagram, the SectorCarrier reference referes to the SectorEquipmentFunction, but is referred to by the EUtranCellxxx MO. When the SectorEquipmentFunction MO is unlocked, the RBS allocates and configures all hardware connected to the sector. It performs the uplink and downlink gain calibration of the antenna branches. When the SectorEquipmentFunction MO is created (as a part of the Site Equipment configuration), it is associated with one or several RfBranch MOs and the parameter settings are persistently stored in the RBS. ManagedElement EnodeBFunction Rcs SecurityHandling EUtraNetwork Paging EUtranCellFDD or EUtranCellTDD SectorCarrier SectorEquipmentFunction SectorEquipmentFunction Has parameters like - Administrative state - Operational state (readonly) - Sector maximum output power - Sector frequency band (readonly) - Reference to the rfbranch This MO is created as a part of the Site Equipment Configuration Figure 2-18: Sector Equipment The following are SectorEquipmentFunction attributes: administrativestate / availabilitystatus /operationalstate States of the sector Ericsson AB 2012 LZT R1A

61 Power Control, Scheduling and Link Adaptation confoutputpower The requested maximum sector power. The value represents the sum of the power for all antenna connectors used by the sector. Note: The output power may be limited by the radio unit hardware and output power licenses. fqband (readonly) The frequency band used in the SectorEquipmentFunction MO, based on the radio that is connected to the MO. reservedby Contains a list of SectorCarrier or EUtranCellFDD MO instances that reserve this MO instance. A maximum of four sector carriers may reserve the same SectorEquipmentFunction MO. rfbranchref Reference to the RfBranch MO instances being reserved. sectorpower Available sector powersecurity handling configuration The values of the attributes of the SecurityHandling MO class determine the security handling characteristics of an RBS. This MO is auto-created with default values. ManagedElement EnodeBFunction Rcs SecurityHandling EUtraNetwork Paging EUtranCellFDD or EUtranCellTDD SectorEquipmentFunction UeMeasControl SecurityHandling Used to configure e.g. - Ciphering algorithms - COUNT-C supervision This MO is auto-created, with default values, by the system Figure 2-19: Security Handling LZT R1A Ericsson AB

62 LTE L13 Radio Network Functionality cipheringalgorithms The attribute determines the priority of the algorithms for ciphering UP data messages and RRC messages. It is possible to set attribute values specifying that some algorithms are not used at all. Setting the value of the cipheringalgorithms to EEA0 turns off the UP data and RRC message ciphering. countwrapsupervisionactive The attribute determines if the RBS supervises COUNT-C wrapping. See 3GPP If the value of the attribute countwrapsupervisionactive is set to true, a new encryption key is generated by releasing and re-establishing the UE connection to the RBS when the supervision function indicates that the COUNT- C is close to its end. This ensures that the same COUNT-C value is never used twice as an encryption input. Caution! If the value of the countwrapsupervisionactive attribute is set to false, no COUNT-C supervision is performed. This may result in the same COUNT-C value being used more than once as input for the encryption, which might compromise security. integrityprotectionalgorithms The attribute determines which of the Integrity Protection Algorithms has the highest priority. The default value is that the AES algorithm has the highest priority, but SNOW 3G is also available. Caution! It is possible to set an attribute value that allows use of only one of the Integrity Protection Algorithms. However, this may cause failed handover of any UEs unable to use the selected Integrity Protection Algorithm UE measurement control configuration All UE Measurement Control attributes can be found in the UeMeasControl MO class and control the system behavior regarding UE measurements. This MO is auto-created with default values. The operator can change the attribute values in order to optimize network performance; however it should be noted that the default parameter setting aims for smooth operation of the system. The most important UE measurement parameters are: Best cell Bad coverage UE measurement filter 3GPP TS defines the quantities Reference Symbol Received Power (RSRP) and Reference Symbol Received Quality (RSRQ) Ericsson AB 2012 LZT R1A

63 Power Control, Scheduling and Link Adaptation ManagedElement EnodeBFunction Rcs SecurityHandling EUtraNetwork Paging EUtranCellFDD or EUtranCellTDD SectorEquipmentFunction UeMeasControl UeMeasControl Used to configure e.g. - Cell quality threshold value - Best cell decision configuration - Bad coverage measurement configuration - UE report filtering This MO is auto-created, with default values, by the system Figure 2-20: UE Measurement Control Further information on the uses of UE measurements can be found in CPI documents Intra-LTE Handover and Performance Management. There are more Managed Objects involved, most of them auto-created, related to Measurement Control and reporting as the following figure shows. LZT R1A Ericsson AB

64 LTE L13 Radio Network Functionality ReportConfigEUtra InterFreqLb ReportConfigA1Prim ReportConfigA1Sec ManagedElement EnodeBFunction ReportConfigEUtra IFBestCell ReportConfigEUtra BadCovPrim ReportConfigEUtra BadCovSec ReportConfigA4 ReportConfigA5 EUtranCellFDD or EUtranCellTDD UeMeasControl ReportConfigEUtra BestCell ReportConfigSearch PmUeMeasControl ReportConfigEUtra IntraFreqPm ReportConfigB2 Cdma2000 ReportConfigB1 Geran ReportConfigB2 Geran ReportConfigB1Utra ReportConfigB2Utra Auto-created with default values Figure 2-21: Measurement and Report related MOs UE Measurement Quality Threshold Configurations smeasure The attribute determines the cell quality threshold value at which the UE starts to perform measurements. The default value is the most sensitive value available in the value range. The attribute is included in the UeMeasControl MO class Best Cell Configuration Optimize mobility behavior by adjusting the values of the following attributes that are included in the reportconfigeutrabestcell MO class: hysteresisa3 The attribute determines the hysteresis for A3 events. Setting an appropriate value for the hysteresisa3 attribute prevents rapid back-and-forth handover. reportquantitya Ericsson AB 2012 LZT R1A

65 Power Control, Scheduling and Link Adaptation The attribute determines what the UE includes in measurement reports. Default value is both, which provides both RSRP and RSRQ measurements. Setting the value to sameastriggerquantity only provides reports with RSRP or RSRQ according to what is set for the attribute triggerquantitya3, which reduces load on the UE. triggerquantitya3 The attribute determines which one of the RSRP and RSRQ measurements are used to trigger A3 events. timetotriggera3 The attribute determines the time to trigger value for A3 event. a3offset The offset value for event A3 determines how much stronger a neighboring cell must be to fulfill the bestcell criterion. reportamounta3 The attribute determines the number of reports for periodical reporting. A value of 0 indicates that an unlimited number of reports can be generated. reportintervala3 The attribute determines the interval for event triggered periodical reporting Poor Coverage Measurement Configuration Poor coverage behavior is handled by event A2 UE measurements. Report Configurations for Primary Poor Coverage UE Measurements The report configurations for primary UE Measurements are included in the ReportConfigEUtraBadCovPrim. Optimize report configurations for primary UE Measurements by adjusting the values of the following attributes: a2thresholdrsrpprim The attribute determines the threshold for A2 events if, and only if, the attribute triggerquantitya2prim is set to RSRP. a2thresholdrsrqprim The attribute determines the threshold for A2 events if, and only if, the attribute triggerquantitya2prim is set to RSRQ. LZT R1A Ericsson AB

66 LTE L13 Radio Network Functionality hysteresisa2prim The attribute determines the hysteresis for primary A2 measurements. Setting an appropriate value for the attribute helps prevent triggering A2 events that are not necessary. reportquantitya2prim The attribute determines the UE content in measurement reports. Default value is both, which provides both RSRP and RSRQ measurements. Setting the value to sameastriggerquantity only provides reports with RSRP or RSRQ according to what is set for the attribute triggerquantitya2prim, which reduces load on the UE. timetotriggera2prim The attribute sets the time to trigger primary A2 events. triggerquantitya2prim The attribute determines whether RSRP or RSRQ is used to trigger primary A2 events. The default value is RSRP, but can be set to RSRQ if desired. reportamounta2prim The attribute determines the number of reports for periodical reporting. A value of 0 indicates that an unlimited number of reports can be generated. reportintervala2prim The attribute determines the interval for event triggered periodical reporting. Report Configurations for Secondary Poor Coverage UE Measurements The report configurations for secondary UE Measurements are included in the ReportConfigEUtraBadCovSec. Optimize report configurations for secondary UE Measurements by adjusting the values of the following attributes: a2thresholdrsrpsec The attribute determines the threshold for A2 events if, and only if, the attribute triggerquantitya2sec is set to RSRP. a2thresholdrsrqsec The attribute determines the threshold for A2 events if, and only if, the attribute triggerquantitya2sec is set to RSRQ. hysteresisa2sec Ericsson AB 2012 LZT R1A

67 Power Control, Scheduling and Link Adaptation The attribute determines the hysteresis for secondary A2 measurements. Setting an appropriate value for the attribute helps prevent triggering A2 events that are not necessary. reportquantitya2sec The attribute determines the UE content in measurement reports. Default value is both, which provides both RSRP and RSRQ measurements. Setting the value to sameastriggerquantity only provides reports with RSRP or RSRQ according to what is set for the attribute triggerquantitya2sec, which reduces load on the UE. triggerquantitya2sec The attribute determines whether RSRP or RSRQ is used to trigger secondary A2 events. The default value is RSRQ, but can be set to RSRP if desired. reportamounta2sec The attribute determines the number of reports for periodical reporting. A value of 0 indicates that an unlimited number of reports can be generated. reportintervala2sec The attribute determines the interval for event triggered periodical reporting Event A1 Measurement Configuration Event A1 measurement implies that a bad coverage condition ceases, that is, the serving cell coverage becomes better than a threshold. Primary and secondary reporting configuration MOs refer to the option to use different settings for two simultaneous measurements. Report Configurations for Primary Event A1 UE Measurements The report configurations for primary A1 UE measurements are included in the ReportConfigA1Prim MO Class. Optimize report configurations for primary UE measurements by adjusting the values of the following attributes: a1thresholdrsrpprim The attribute determines the threshold for A1 events if, and only if, the attribute triggerquantitya1prim is set to RSRP. a1thresholdrsrqprim The attribute determines the threshold for A1 events if, and only if, the attribute triggerquantitya1prim is set to RSRQ. hysteresisa1prim LZT R1A Ericsson AB

68 LTE L13 Radio Network Functionality The attribute determines the hysteresis for primary A1 measurements. Setting an appropriate value for the attribute helps prevent triggering A1 events that are not necessary. reportquantitya1prim The attribute determines the UE content in measurement reports. Default value is both, which provides both RSRP and RSRQ measurements. Setting the value to sameastriggerquantity only provides reports with RSRP or RSRQ according to what is set for the attribute triggerquantitya1prim. This reduces load on the UE. timetotriggera1prim The attribute sets the time to trigger primary A1 events. triggerquantitya1prim The attribute determines whether RSRP or RSRQ is used to trigger primary A1 events. The default value is RSRP, but can be set to RSRQ if desired. reportamounta1prim The attribute determines the number of reports for periodical reporting. A value of 0 indicates that an unlimited number of reports can be generated. reportintervala1prim The attribute determines the interval for event triggered periodical reporting. Report Configurations for Secondary Event A1 UE Measurements The report configurations for secondary UE measurements are included in the ReportConfigA1Sec MO Class. Optimize report configurations for secondary UE Measurements by adjusting the values of the following attributes: a1thresholdrsrpsec The attribute determines the threshold for A1 events if, and only if, the attribute triggerquantitya1sec is set to RSRP. a1thresholdrsrqsec The attribute determines the threshold for A1 events if, and only if, the attribute triggerquantitya1sec is set to RSRQ. hysteresisa1sec The attribute determines the hysteresis for secondary A1 measurements. Setting an appropriate value for the attribute helps prevent triggering A1 events that are not necessary Ericsson AB 2012 LZT R1A

69 Power Control, Scheduling and Link Adaptation reportquantitya1sec The attribute determines the UE content in measurement reports. Default value is both, which provides both RSRP and RSRQ measurements. Setting the value to sameastriggerquantity only provides reports with RSRP or RSRQ according to what is set for the attribute triggerquantitya1sec. This reduces load on the UE. triggerquantitya1sec The attribute determines whether RSRP or RSRQ is used to trigger secondary A1 events. The default value is RSRQ, but can be set to RSRP if desired. reportamounta1sec The attribute determines the number of reports for periodical reporting. A value of 0 indicates that an unlimited number of reports can be generated. reportintervala1sec The attribute determines the interval for event triggered periodical reporting Event A5 Measurement Configuration The A5 event implies that the serving cell becomes worse than a first threshold, and neighbor cell becomes better than a second threshold. The report configurations for event A5 UE measurements are included in the ReportConfigA5 MO Class. Optimize report configurations for event A5 UE measurements by adjusting the values of the following attributes: a5threshold1rsrp The attribute determines the first threshold value for A5 events if, and only if, the attribute triggerquantitya5 is set to RSRP. a5threshold2rsrp The attribute determines the second threshold value for A5 events if, and only if, the attribute triggerquantitya5 is set to RSRP. a5threshold1rsrq The attribute determines the first threshold value for A5 events if, and only if, the attribute triggerquantitya5 is set to RSRQ. a5threshold2rsrq The attribute determines the second threshold value for A5 events if, and only if, the attribute triggerquantitya5 is set to RSRQ. LZT R1A Ericsson AB

70 LTE L13 Radio Network Functionality hysteresisa5 The attribute determines the hysteresis for A5 measurements. Setting an appropriate value for the attribute helps prevent triggering A5 events that are not necessary. reportquantitya5 The attribute determines the UE content in measurement reports. Default value is both, which provides both RSRP and RSRQ measurements. Setting the value to sameastriggerquantity only provides reports with RSRP or RSRQ according to what is set for the attribute triggerquantitya5. This reduces load on the UE. timetotriggera5 The attribute sets the time to trigger A5 events. triggerquantitya5 The attribute determines whether RSRP or RSRQ is used to trigger primary A5 events. The default value is RSRP, but can be set to RSRQ if desired. reportamounta5 The attribute determines the number of reports for periodical reporting. A value of 0 indicates that an unlimited number of reports can be generated. reportintervala5 The attribute determines the interval for event triggered periodical reporting. maxreportcellsa5 The attribute determines the maximum number of cells to include in the measurement report for the event A5 measurement Event B2 Measurement Configuration The B2 event implies that the serving cell becomes worse than a first threshold, and an interrat neighbor cell becomes better than a second threshold. This measurement confirms if the UE listens to cells (UTRA FDD and CDMA2000 ehrpd) or frequencies (GERAN), before a release with redirect is triggered to those technologies. The report configurations for event B2 measurements are included in MO Classes ReportConfigB2Utra, ReportConfigB2Cdma2000 and ReportConfigB2Geran respectively. Event B2 Report Configurations for UTRA Measurements Ericsson AB 2012 LZT R1A

71 Power Control, Scheduling and Link Adaptation The MO Class ReportConfigB2Utra contains report configuration settings for the B2 event for UTRA measurements. The event is used to detect when the serving cell becomes worse than a first threshold and a UTRAN cell becomes better than a second threshold. Optimize report configurations for UTRA Measurements by adjusting the values of the following attributes: b2threshold1rsrp The attribute determines the first threshold for B2 events if, and only if, the attribute triggerquantityb2 is set to RSRP. b2threshold1rsrq The attribute determines the first threshold for B2 events if, and only if, the attribute triggerquantityb2 is set to RSRQ. b2threshold2rscputra The attribute determines the second threshold (UTRA threshold) for B2 events if, and only if, the attribute triggerquantityb2 is set to RSCP. b2threshold2ecnoutra The attribute determines the ratio of energy per modulation bit to noise spectral density in the second threshold value (UTRA threshold) for B2 events. This only applies if, and only if, the attribute triggerquantityb2 is set to ECNO. This parameter is not applicable for UTRA TDD. hysteresisb2 The attribute determines the hysteresis for B2 event measurements on EUTRAN. Setting an appropriate value for the attribute helps prevent triggering B2 events that are not necessary. timetotriggerb2 The attribute sets the time to trigger B2 events. triggerquantityb2 The attribute determines the quantity for the first threshold that is sent to the UE, and is used together with the second threshold to trigger the event B2. reportamountb2 The attribute determines the number of reports for periodical reporting. A value of 0 indicates that an unlimited number of reports can be generated. reportintervalb2 LZT R1A Ericsson AB

72 LTE L13 Radio Network Functionality The attribute determines the interval for event triggered periodical reporting. maxreportcellsb2 The attribute determines the maximum number of cells to include in the measurement report for the event B2 measurement. Event B2 Report Configurations for CDMA2000 Measurements The MO Class ReportConfigB2Cdma2000 contains report configuration settings for the B2 event for CDMA2000 measurements. The event is used to detect when the serving cell becomes worse than a first threshold, and a CDMA2000 cell becomes better than a second threshold. Optimize report configurations for CDMA2000 measurements by adjusting the values of the following attributes: b2threshold1rsrp The attribute determines the first threshold for B2 events if, and only if, the attribute triggerquantityb2 is set to RSRP. b2threshold1rsrq The attribute determines the first threshold for B2 events if, and only if, the attribute triggerquantityb2 is set to RSRQ. b2threshold2cdma2000 The attribute determines the signal quality in the second threshold (CDMA2000 threshold), for B2 events. hysteresisb2 The attribute determines the hysteresis for B2 event measurements on EUTRAN. Setting an appropriate value for the attribute helps prevent triggering B2 events that are not necessary. timetotriggerb2 The attribute sets the time to trigger B2 events. triggerquantityb2 The attribute determines the quantity for the first threshold that is sent to the UE, and is used together with the second threshold to trigger the event B2. reportamountb2 The attribute determines the number of reports for periodical reporting. A value of 0 indicates that an unlimited number of reports can be generated. reportintervalb Ericsson AB 2012 LZT R1A

73 Power Control, Scheduling and Link Adaptation The attribute determines the interval for event triggered periodical reporting. maxreportcellsb2 The attribute determines the maximum number of cells to include in the measurement report for the event B2 measurement. Event B2 Report Configurations for GERAN Measurements The MO Class ReportConfigB2Geran contains report configuration settings for the B2 event for GERAN measurements. The event is used to detect when a GERAN cell becomes better than a second threshold and the serving cell becomes worse than a first threshold. Optimize report configurations for GERAN measurements by adjusting the values of the following attributes: b2threshold1rsrp The attribute determines the first threshold for B2 events if, and only if, the attribute triggerquantityb2 is set to RSRP. b2threshold1rsrq The attribute determines the first threshold for B2 events if, and only if, the attribute triggerquantityb2 is set to RSRQ. b2threshold2geran The attribute determines the signal quality in the second threshold (GERAN threshold), for B2 events. hysteresisb2 The attribute determines the hysteresis for B2 event measurements on EUTRAN. Setting an appropriate value for the attribute helps prevent triggering B2 events that are not necessary. timetotriggerb2 The attribute sets the time to trigger B2 events. triggerquantityb2 The attribute determines the quantity for the first threshold that is sent to the UE, and is used together with the second threshold to trigger the event B2. reportamountb2 The attribute determines the number of reports for periodical reporting. A value of 0 indicates that an unlimited number of reports can be generated. LZT R1A Ericsson AB

74 LTE L13 Radio Network Functionality reportintervalb2 The attribute determines the interval for event triggered periodical reporting. maxreportcellsb2 The attribute determines the maximum number of cells to include in the measurement report for the event B2 measurement UE Measurement Filter Configuration Filtering is used to improve event evaluation based on UE measurements. It helps prevent isolated anomalous UE measurements from influencing event evaluation too much. The attributes are included in the UeMeasControl MO Class. The following filtering coefficients exist: filtercoefficienteutrarsrp The attribute determines the filtering coefficient for EUTRA using measured RSRP. filtercoefficienteutrarsrq The attribute determines the filtering coefficient for EUTRA using measured RSRQ. For further information, see the chapter about Layer 3 Filtering in 3GPP TS Event Based ANR Measurements Event A3 for ANR The purpose with event based ANR measurements is to detect neighbor cells before UE have reported them in mobility reports for handover. The process employed by the UE for the event based evaluations of surrounding cells uses parameters sent by the serving RBS to the UE. The ANR function use reports from the UE to identify neighbor cells. For parameters to adjust the report volume, see CPI ANR Function. The A3 event implies that one or several neighbor cells become better than the serving cell also when some offset and hysteresis values are taken into account. Event A3 for ANR use the following parameters in the ReportConfigEUtraBestCellAnr MO class, relating to the general event A3 parameters: a3offsetanrdelta Ericsson AB 2012 LZT R1A

75 Power Control, Scheduling and Link Adaptation hysteresisa3 timetotriggera3 Event A5 for ANR The A5 event implies that the serving cell becomes worse than a first threshold, and neighbor cell becomes better than a second threshold. Event A5 for ANR use the following parameters relating to the general event A5 parameters in the ReportConfigA5 MO Class: triggerquantitya5 a5threshold1rsrpanrdelta a5threshold1rsrqanrdelta a5threshold2rsrpanrdelta a5threshold2rsrqanrdelta hysteresisa5 timetotriggera5 LZT R1A Ericsson AB

76 LTE L13 Radio Network Functionality 6 QoS Configuration QoS Class Identifiers (QCI) parameters are sent in a call setup to the RBS from the Core Network. Each Data Radio Bearer (DRB) that is set up is mapped to a specific QCI value. The QCI table in the RBS6000 is used to map the incoming parameters to QoS settings for the Radio Network (e.g. scheduling) and Transport Network (DSCP values). Traffic separation is a prerequisite for QCI-differentiated prioritization mechanisms to effectively act on bearers in enb. Non-standardized QCIs are all given the same priority, which shall be lower compared to priorities for the standardized QCIs. For the uplink, the priorities are sent to the UE, which may differentiate/prioritize between its logical channels. Mapping QCIs to Logical Channel Groups (LCGs) can be configured as described later in this section, and it enables traffic separation in the uplink. There are three LCGs (1-3) available. By default, LCG 1 is assigned to all QCIs. Mapping QCI to DiffServ Code Point (DSCP) for the uplink over S1 and in the downlink for packet forwarding over X2 can be configured as also is described later in this section. The DSCP setting determines the priority for the data stream in the IP transport network. Several QCIs can be mapped to the same DSCP value. Non-standardized QCIs are all given the same configurable DSCP value. For the uplink (enodeb towards the EPC) traffic, the transport network benefits from QoS by mapping QCI to DSCP in the RBS. This enables the transport network to prioritize between its different data flows over the S1 interface in the uplink and over the X2 interface for the downlink data in case of Packet Forwarding. For the DL, a similar mapping is performed in the S-GW for the S1 DL data. All QoS class identifiers defined by 3GPP are accepted, including the ones that are outside the standardized range QCI1-9. Note that QCI / LCG /DSCP mentioned in this section is only a part of the total QoS concept in the enodeb. Also the configurations in the enodeb are just a part of the total QoS implementation. Only if UE, enodeb, Transport Network and Packet Data Network are configured correctly could the performance of different services be ensured Radio Bearer and QoS Managed Objects This section describes the configurations available in the Radio Bearer part of the ENodeBFunction Managed Area (MA). The mapping for a specific QCI entry to a DSCP value can be changed by editing some parameters (see figure below). It should also be noted that the default values of the QCI-related attributes generally ensure smooth operation, but it is possible for the operator to change some attributes to optimize performance Ericsson AB 2012 LZT R1A

77 Power Control, Scheduling and Link Adaptation QCI relates to the DRB; consequently, the Signaling Radio Bearer (SRB) is not affected by QciTable MOs. QCI 1 Prio 2 LCG 1 DSCP 46 Default config : : : : : : : : ManagedElement 9 0, EnodeBFunction One per QCI entry (QCI 0 and are called default and have lower prio than 1-9) The attribute dscp is changed in order to map the QCI value to a new DSCP value. QosProfilePredefined 1..1 QciTable LogicalChannelGroup Uplink traffic separation is enabled with Logical Channel Groups. EUtranCellFDD or EUtranCellTDD These MOs are auto-created, with default values, by the system. May need to be changed for a different QoS handling. Figure 2-22: MOs related to QoS configuration The dscp attribute of the QciProfilePredefined MO defines the mapping between QCI and Differentiated Services Code Point (DSCP). This corresponds to mapping from RAN to Transport Network. The operator's view of the QCI to DSCP mapping may differ from that of the default values. If so, the operator can change the value of the dscp attributes to better reflect that view of the relationship between QCI and DSCP. The UE sends Buffer Status Reports (BSRs) to the enodeb, stating how much data it has in its buffers (data to be sent in uplink). If the UE uses several Logical Channels, there would, in theory, be separate BSRs per channel. To simplify for the Scheduler in the enodeb and to minimize the amount of signaling traffic sent in the uplink, logical channels can be grouped into up to three different Logical Channel Groups (LCGs). There is only one BSR sent per LCG, and the Scheduler bases its decision per LCG. The logicalchannelgroupref attribute of the QciProfilePredefined MO assigns an LCG for a QCI. Each QCI must be assigned to one of the three LCGs. An LCG may have several QCIs assigned to it. The IP DiffServ field and Ethernet Pbits support various Quality of Service (QoS) levels for the user and control plane traffic in the WCDMA RAN. This allows delay-sensitive traffic to be given priority over less sensitive traffic types in times of congestion. LZT R1A Ericsson AB

78 LTE L13 Radio Network Functionality Transport Network Qos concepts Differentiated Services Code Point (DSCP) In the IPv4 header, there is one octet (8-bits) assigned as the Type Of Service (ToS)or also called the DiffServ field. The six most significant bits of the DiffServ field are called the Differentiated Services Code Point (DSCP) while the last two are used as Explicit Congestion Notification (ECN) bits allowing advance notification of congestion. The RAN nodes tag the egress (outgoing) IP packets with a configured DSCP value depending on the RAB type according to RFC 2474 and 2475 Internet standards. Other network devices in the network that support Diffserv use the DSCP value in the IP header to select the Per Hop Basis (PHB) behavior for the packet and provide the appropriate QoS treatment. The routing device prioritizes traffic by class first based on the Precedence Level specified by the 3 most significant bits of the DiffServ field (DS5-DS3). Then it differentiates and prioritizes same-class traffic, taking the drop probability into account-based on the next three bits (DS2-DS0). The Operator can configure the DSCP values for the S1 and X2 interfaces and then use the QCI (QoS Class Identifier) at call setup to map a specific user to a DSCP value as shown below Ethernet Pbits Ethernet Switches offer a different QoS level to each Ethernet frame depending on the value of the Pbit field. The mapping of Pbit to physical port QoS Queue is automatically done in the socalled ET-IP hardware in an RBS6000 (the ET-IP is the part where the physical IP interfaces are located in the DUL). The ET-IP is modeled by a logical board described by the MO ExchangeTerminalIp. The mapping from Pbit to physical port QoS Queue is done in eight levels. Other vendor equipment used in the backbone may support just two or maybe four priority queues in their Ethernet switches. It is therefore of utmost importance to understand how different Ethernet vendors have mapped the different priority levels to the device's traffic queues. If the mapping is done wrong in only one node in the network, the whole QoS implementation may fail resulting in severe congestion. The scheduler schedules packets coming from the Packet Data Convergence Protocol (PDCP) layer in a strict priority order, with the highest queue number being served first (8->7->6->5->4->3->2->1). Thus, if there are packets in a higher prioritized queue, those are always sent before packets in lower prioritized queues. With this scheduling mechanism, there is a risk that lower prioritized queues are starved of bandwidth when the traffic approaches the limit of the Ethernet Gigabit port. With proper link dimensioning the likelihood of this happening is diminished. After passing through the Strict Priority Scheduler the DSCP-to-Pbit-mapping is performed Ericsson AB 2012 LZT R1A

79 Power Control, Scheduling and Link Adaptation 7 Summary Configure the Radio Network in RBS6000 Explain the concept of cell and its relation to sector and antennae system in RBS6000 Recognize the Managed Objects related to radio network configuration Identify some basic parameters related to cell and cell relations Identify, and, if necessary, change QoS related parameters in RBS6000 Figure 2-23: Summary of Chapter 2 LZT R1A Ericsson AB

80 LTE L13 Radio Network Functionality Intentionally Blank Ericsson AB 2012 LZT R1A

81 Radio Connection Supervision 3 Idle Mode Behavior Objectives After this chapter the participants will be able to: Explain the purpose and function of Idle Mode Behavior 1. Explain PLMN and Cell selection and reselection 2. Explain registration updating procedures 3. Explain paging procedures 4. Describe the organization of the System Information Figure 3-1: Objectives of Chapter 3 LZT R1A Ericsson AB

82 LTE L13 Radio Network Functionality 1 Introduction In Idle mode the User Equipment (UE) has no connection to the radio network, i.e. no Radio Resource Control (RRC) connection is established. The purpose of keeping UEs in Idle mode is to minimize the resource usage both for the UEs and for the network. Yet the UEs should still be able to access the system and be reached by the system with acceptable delays. In Idle mode, the UE can move in the whole area covered by cells belonging to a PLMN and still be able to set up or receive calls from the network. In order to keep the paging load on a reasonable level, the cells belonging to a PLMN are divided into Tracking Areas (TAs). The UE makes its presence known to the CN when moving into a new TA (-list) by performing a registration in the new area (TA update). The UE also typically makes its presence known at power on and indicates that it is not reachable at power off. For this purpose the attach/detach procedure is used. This procedure makes it possible for the CN to avoid unnecessary paging attempts. Reading System Information PLMN selection and reselection Cell selection and reselection TA Update Paging procedure Figure 3-2. Idle Mode tasks Idle mode behavior is managed by the system information that is sent on the Broadcast Control Channel (BCCH) in each cell. The system information contains parameters that control cell selection and reselection, paging, TA update, access and also parameters related to other functions. By using different parameter settings in system information, the operator is allowed to modify the service area and the cells that the UE can camp on Ericsson AB 2012 LZT R1A

83 Radio Connection Supervision The Idle mode tasks may be divided into five different processes: Reading System Information PLMN selection and reselection Cell selection and reselection Tracking area (TA) update Paging procedure Figure 3-3 illustrates the relationship between PLMN selection, cell selection and reselection and TA update. Note: Closed Subscriber Group (CSG) Selection not in L12. TS Manual Mode PLMN Selection Automatic mode CSG ID selection Indication to user Location Registration response PLMNs available Support for manual CSG ID Selection AvailableCSG IDs to NAS PLMN selected Registration Area changes CSG ID selected Cell Selection and Reselection NAS Control Radio measurements Service requests Location Registration Figure 3-3 The relationship between some of the idle mode tasks. The UE may, if necessary, register its presence, by means of a NAS registration procedure, in the tracking area of the chosen cell and as outcome of a successful Location Registration the selected PLMN becomes the registered PLMN. If the UE finds a more suitable cell, according to the cell reselection criteria, it reselects onto that cell and camps on it. If the new cell does not belong to at least one tracking area to which the UE is registered, location registration is performed (typically using the TA update procedure). If necessary, the UE may search for higher priority PLMNs at regular time intervals and search for a suitable cell if another PLMN has been selected by NAS. LZT R1A Ericsson AB

84 LTE L13 Radio Network Functionality If the UE loses coverage of the registered PLMN, either a new PLMN is selected automatically (automatic mode), or an indication of which PLMNs are available is given to the user, so that a manual selection can be made (manual mode). The purpose of camping on a cell in idle mode is fourfold: a) It enables the UE to receive system information from the PLMN. b) When registered and if the UE wishes to establish an RRC connection, it can do this by initially accessing the network on the control channel of the cell on which it is camped. c) If the PLMN receives a call for the registered UE, it knows (in most cases) the set of tracking areas in which the UE is camped. It can then send a "paging" message for the UE on the control channels of all the cells in this set of tracking areas. The UE then receives the paging message because it is tuned to the control channel of a cell in one of the registered tracking areas and the UE can respond on that control channel. d) It enables the UE to receive ETWS (Earth quake and Tsunami Warning System) notifications (not supported in L12). If the UE is unable to find a suitable cell to camp on, or the USIM is not inserted, or if the location registration failed, the UE enters the "limited service" state in which it can only attempt to make emergency calls. E-UTRAN/EPC in Rel-8 does not support USIM-less emergency calls. A UE not equipped with a valid USIM (i.e. no USIM or SIM only) will disable all its E- UTRAN capabilities In Idle mode, the UE selects a suitable cell of the selected Public Land Mobile Network (PLMN) to camp on by using the cell selection algorithm. After the cell selection the UE attaches and registers to the Core Network (CN) supported by the PLMN. In Idle mode the UE monitors system and paging information. The UE also performs cell reselection based on radio measurements. Cell reselection makes sure that the UE is always camping on the cell that gives the highest probability for successful monitoring of system and paging information and for successful establishment of a connection. The cell reselection process may imply a change of the Radio Access Technology (RAT) i.e. (GSM/ GPRS /WCDMA/CDMA2000 LTE). 2 System Information System information distribution is an essential and basic feature of the LTE RAN. The UE requires System Information to: Perform idle mode tasks PLMN selection Cell selection and reselection Conform to Cell Reservations and Access Restrictions Ericsson AB 2012 LZT R1A

85 Radio Connection Supervision Tracking Area registration Originate access to the network Terminate access including reception of paging messages Maintain its LTE connection Perform handover 2.1 Broadcast of system information System information is divided into the Master Information Block (MIB) and a number of System Information Blocks (SIBs). SIB1 Parameters SIB2: related to PLMN & Cell Access SIB3 Parameters plmn-identitylist related to Cell Selection ac-barringinfo trackingareacode SIB4 cellreselectioninfocommon radioresourceconfigcommon cellidentity RadioResourceConfigCommonSIB, intrafreqneighcelllist cellbarred q-hyst intrafreqreselectio speedstatereselectionpars ue-timersandconstants intrafreqblackcelllist csg-indication mobilitystateparameters freqinfo csg-physcellidrange csg-identity q-hystsf {sf-medium, sf-high} Parameters cellreselectionservingfreqinfo ul-carrierfreq related to Cell Selection ul-bandwidth PhysCellIdRange q-rxlevmin s-nonintrasearch {n6, n15, n25, n50, n75, n100} additionalspectrumemission physcellid q-rxlevminoffset threshservinglow p-maxcellreselectionpriority IntraFreqNeighCellInfo freqbandindicator SIB11: q-offsetcell intrafreqcellreselectioninfo schedulinginfolist SIB10: tdd-config q-rxlevmin SIB9: OffsetRange WindowLength p-max systeminfovaluetag s-intrasearch SIB8: allowedmeasbandwidth SIB7: presenceantennaport1 SIB6: neighcellconfig, SIB5: t-reselectioneutra, SIB4: SIB3: SIB2: SIB1: MIB: DL BW Figure 3-4: System Information Broadcast SI - RNTI PBCH: MIB DLSCH:SIBs PSC+SSC (Physical Cell Id) PSC+SSC (Physical Cell Id) SIB The main idea with SIBs is to group system information elements of the same nature, for example, cell selection and cell reselection parameters. -MasterInformationBlock defines the most essential physical layer information of the cell required to receive further system information; -SystemInformationBlockType1 contains information relevant when evaluating if a UE is allowed to access a cell and defines the scheduling of other system information blocks. SIB1 Parameters: Parameters related to PLMN & Cell Access plmn-identitylist (plmn-identity, cellreservedforoperatoruse) trackingareacode cellidentity cellbarred intrafreqreselectio LZT R1A Ericsson AB

86 LTE L13 Radio Network Functionality csg-indication Parameters related to Cell Selection q-rxlevmin q-rxlevminoffset p-max freqbandindicator schedulinginfolist (si-periodicity, sib-mappinginfo) tdd-config si-windowlength systeminfovaluetag -SystemInformationBlockType2 contains common and shared channel information SIB2 Parameters: Parameters related to Cell Selection ac-barringinfo radioresourceconfigcommon(common ch and paging parameters) ue-timersandconstants ul-bandwidth {n6, n15, n25, n50, n75, n100} additionalspectrumemission timealignmenttimercommon -SystemInformationBlockType3 contains cell re-selection information, mainly related to the serving cell. SIB3 Parameters: Parameters related to Cell Reselection q-hyst q-hystsf {sf-medium, sf-high} snonintrasearch threshservinglow cellreselectionpriority qrxlevmin p-max sintrasearch allowedmeasbandwidth presenceantennaport1 neighcellconfig t-reselectioneutra t-reselectioneutra-sf -SystemInformationBlockType4 contains information about the serving frequency and intra-frequency neighboring cells relevant for cell re-selection (including cell re-selection parameters common for a frequency as well as cell specific reselection parameters) SIB4 Parameters: Ericsson AB 2012 LZT R1A

87 Radio Connection Supervision intrafreqneighcelllist (physcellid, q-offsetcell) q-offsetcell intrafreqblackcelllist (physcellidrange) csg-physcellidrange -SystemInformationBlockType5 contains information about other E-UTRA frequencies and inter-frequency neighboring cells relevant for cell re-selection (including cell re-selection parameters common for a frequency as well as cell specific re-selection parameters); SIB5 Parameters: InterFreqCarrierFreqList InterFreqCarrierFreqInfo dl-carrierfreq q-rxlevmin p-max t-reselectioneutra t-reselectioneutra-sf threshx-high threshx-low allowedmeasbandwidth presenceantennaport1 cellreselectionpriority neighcellconfig q-offsetfreq interfreqneighcelllist interfreqblackcelllist InterFreqNeighCellInfo physcellid q-offsetcell InterFreqNeighCellList InterFreqBlackCellList (physcellidrange) -SystemInformationBlockType6 contains information about UTRA frequencies and UTRA neighboring cells relevant for cell re-selection (including cell reselection parameters common for a frequency as well as cell specific re-selection parameters); SIB6 Parameters: t-reselectionutra t-reselectionutra-sf carrierfreqlistutra-fdd (carrierfreq, cellreselectionpriority, threshx-high, threshx-low, qrxlevmin, p-maxutra, q- QualMin) LZT R1A Ericsson AB

88 LTE L13 Radio Network Functionality carrierfreqlistutra-tdd threshx-high threshx-low q-rxlevmin p-maxutra q-qualmin threshx-highq threshx-lowq -SystemInformationBlockType7 contains information about GERAN frequencies relevant for cell re-selection (including cell re-selection parameters for each frequency); SIB7 Parameters: carrierfreqsinfolist carrierfrequencyinfolistgeran (carrierfreqs, commoninfo, cellreselectionpriority, ncc-permitted, qrxlevmin, p-maxgeran, threshx-high, threshx-low) commoninfo t-reselectiongeran t-reselectiongeran-sf -SystemInformationBlockType8 contains information about CDMA2000 frequencies and CDMA2000 neighboring cells relevant for cell re-selection (including cell re-selection parameters common for a frequency as well as cell specific re-selection parameters); SIB8 Parameters: systemtimeinfo searchwindowsize parametershrpd (preregistrationinfohrpd, cellreselectionparametershrpd) parameters1xrtt (csfb-supportfordualrxues-r9) neighcellcdma2000(bandclass, neighcellsperfreqlist) Ericsson AB 2012 LZT R1A

89 Radio Connection Supervision neighcellsperbandclasscdma2000 (arfcn, physcellidcdma2000) cellreselectionpriority threshx-high threshx-low t-reselectioncdma2000 t-reselectioncdma2000-sf -SystemInformationBlockType9 contains a home enb identifier (HNBID); -SystemInformationBlockType10/11 contain an ETWS primary/secondary notification; Note: SIB 9, 10, 11 not supported in L12. The MIB is mapped on the BCCH and carried on BCH while all other SI messages are mapped on the BCCH and dynamically carried on DL-SCH where they can be identified through the SI-RNTI (System Information RNTI). SIB5 InterFreqCarrierFreqList f 1 SIB5 E-UTRA SIB f 2 InterFreqCarrierFreqInfo dl-carrierfreq q-rxlevmin p-max t-reselectioneutra t-reselectioneutra-sf threshx-high threshx-low allowedmeasbandwidth presenceantennaport1 cellreselectionpriority neighcellconfig q-offsetfreq interfreqneighcelllist interfreqblackcelllist enodeb Inter-frequency enodeb InterFreqNeighCellInfo physcellid q-offsetcell InterFreqNeighCellList InterFreqBlackCellList Figure 3-5 System Information: Inter-frequency mobility. LZT R1A Ericsson AB

90 LTE L13 Radio Network Functionality The redirection priorities set per cell: connectedmodemobilityprio The priorities can be set for : GeranFreqGroup UtranFrequency Cdma2000Freq enodeb IRAT SIB SIB6 - UTRA SIB7 - GERAN SIB8 CDMA2000 SIB8 treselectioncdmahrpd SIB7 treselectioncdmahrpdsfhigh treselectiongran SIB6 treselectioncdmahrpdsfmedium treselectiongransfhigh treselectionutra hrpdcellreselectionpriority treselectiongransfmedium treselectionutrasfhigh hrpdbandclass cellreselectionpriority treselectionutrasfmedium pmaxgeran qrxlevmin treshxhighhrpd pmaxgeran treshxlowhrpd treshx-high freqcdma treshx-low pnoffset Figure 3-6 System Information IRAT Mobility: Measurement Configuration (GSM, WCDMA and CDMA2000) Note that connectedmodemobilityprio is used for RRC_Connected mode. However it sends the UE, in RRC_Connected, to a chosen RAT in Idle Mode. LTE parameters that define System Information can be found in the MOM according to the Figure 3-7. Managed Object: EUtranCellFDD EUtranCellFDD SIB1, SIB2 parameters SIB3 SIB4 SIB5 SIB6 SIB7 SIB8 Figure 3-7 System Information Configuration Ericsson AB 2012 LZT R1A

91 Radio Connection Supervision 3 PLMN SELECTION PLMN selection is the first step in the registration process that allows a UE to carry out or receive services from an operator. The PLMN may be the Home PLMN (HPLMN), or another PLMN. The UE normally operates on its HPLMN. However, a Visited PLMN (VPLMN) may be selected if the UE loses coverage or e.g. goes abroad. A UE successfully registers on a PLMN if it finds a suitable cell to camp on within the selected PLMN and obtains a location or routing registration acknowledgement in the area of the cell on which it is camped. The UE displays to the user that this PLMN is registered. When a UE does not find a suitable cell in the selected PLMN, it tries to camp on any other acceptable cell within an allowed PLMN. A cell is suitable if it belongs to the selected PLMN, the cell selection criteria (see section Cell selection procedure ) are fulfilled, the cell is not barred and if it is not part of a forbidden TA. This leads to the Camped Normally state. An acceptable cell is a cell belonging to any PLMN or TA, the cell selection criteria are fulfilled and the cell is not barred. This leads to the Camped on any cell state. In this limited service state, the UE is able only of making emergency calls. In this state, the UE repeatedly attempts to find a suitable cell. It scans all radio frequencies (intra-, inter and IRAT frequencies, according to its capabilities). A PLMN is available when at least one acceptable cell for that PLMN exists. When there is no available PLMN, the UE displays the no service state to the user. The user may choose between two different modes of PLMN selection: Automatic or Manual mode. The enodeb includes one PLMN identity in the plmn-identitylist in the System Information Block (SIB) type 1 in each cell, indicating the PLMN identity to which the cell belongs. 3.1 Automatic mode If automatic mode is chosen, the PLMN selection can follow different procedures as described in the following. LZT R1A Ericsson AB

92 Power Spectrum Magnitude (db) Power Spectrum Magnitude (db) Power Spectrum Magnitude (db) Frequency Frequency LTE L13 Radio Network Functionality Stored information for cell selection RSRP 4. USIM x 10 7 RSRP 3. BCH: System info f 1 f n RSRP x 10 7 Figure 3-8 Stored information for cell selection in automatic mode. 1 When the UE is switched on it checks if it is possible to camp on the last registered PLMN. To speed up the PLMN selection procedure the UE uses information about the last registered PLMN such as carrier frequency. 2 The UE searches for the cell with the strongest signal. 3 The UE reads the system information for the PLMN identity (mnc and mcc) on the strongest signal cell. If the UE finds a suitable cell, it tries to register. 4 If the registration is not successful, the UE scans all the carrier frequencies that are stored on the USIM (repeat step 2-3) Ericsson AB 2012 LZT R1A

93 Power Spectrum Magnitude (db) Power Spectrum Magnitude (db) Power Spectrum Magnitude (db) Radio Connection Supervision Initial Cell Selection 5. f 1 Strongest cell Frequency x 10 7 I II PLMN USIM HPLMN USIM PLMN PLMN I II III 6. PLMN A B C D E F III PLMN USIM PLMN PLMN I II III Hz IV PLMN PLMN PLMN Frequency x 10 7 PLMN A PLMN B PLMN D PLMN E PLMN B PLMN E PLMN D PLMN A V PLMN PLMN PLMN Frequency x 10 7 Figure 3-9 Initial cell selection in automatic mode. If the UE does not find a suitable cell on the last registered PLMN it: 5 Searches for the cell with the strongest signal in order to find an available PLMN on a carrier frequency. 6 Scans all RF frequencies (repeat step 5). 7 The available PLMNs are put in a list in priority order. The UE selects the highest ranked PLMN and tries to register. The priority order is: I. Home PLMN, if not previously selected, according to the Radio Access Technologies (RATs) supported by the UE. If an EHPLMN (Equivalent HPLMN) list is present, the highest priority EHPLMN that is available is chosen. II. Each PLMN in the user-controlled PLMN list in the USIM, if present, in order of priority, according to the RATs supported by the UE. III. Each PLMN in the operator-controlled PLMN list in the USIM, in order of priority, according to the RATs supported by the UE. LZT R1A Ericsson AB

94 Power Spectrum Magnitude (db) LTE L13 Radio Network Functionality IV. Other PLMNs, according to the high-quality criterion, in random order. A PLMN is considered to be a high-quality if the Reference Symbol Received Power (RSRP) is dbm. For GSM cells the highquality criterion is fulfilled when the signal level is above 85dBm and for WCDMA above or equal to -95dBm. V. Other PLMNs, in order of decreasing signal strength. If it is a multi-mode UE, it scans all carrier frequencies, according to the UEsupported RATs, in order to select a PLMN to register on. If registration is successful, the UE displays the selected PLMN. The UE may also find more than one available and allowed PLMN, but it may fail registration on those PLMNs if it enters a barred cell or barred TA before the registration attempt. In this case, the UE selects the first one of these PLMNs and enters a limited service state. If it does not find an available PLMN, the UE enters the non-service state, and waits until a new PLMN is available. 3.2 Manual mode The user may also choose manual mode to select a PLMN according to the following procedure, Figure Strongest cell Frequency x 10 7 f 1 f 2 PLMN A PLMN B PLMN D PLMN E 10. f n Figure 3-10 PLMN selection in manual mode. 8 The UE scans all RF channels and searches for the strongest cell signal on each carrier. 9 The UE displays those PLMNs that are allowed as well as those that are not allowed based on the strongest signal cell on each frequency. 10 The user can select a PLMN manually from the list Ericsson AB 2012 LZT R1A

95 Radio Connection Supervision 3.3 Roaming Roaming is a service through which a UE in a given PLMN is able to obtain services from another PLMN in the same country (national roaming) or another country (international roaming). The behavior that the UE has to follow is specified by agreements among the network operators. If the UE in Automatic mode has selected and registered on an allowed VPLMN in its home country, the UE periodically makes attempts to return to its HPLMN. The time interval between two consecutive attempts is stored in the USIM and is managed by the network operator using a timer. The timer can have a value of between 6 minutes and 8 hours, with a step size of 6 minutes. In the absence of a fixed value, a default value of 60 minutes is used by the UE. 4 Network Cell Access restriction The action of camping on a cell is necessary for the UE to get access to some services in the network. Three types of service are defined for the UE in idle mode: Figure Defined Services for the UE in Idle Mode: Normal Services Normal services, for public use on a suitable cell. Important Related Attributes: activeplmnlist Lists the PLMN IDs served by at least one MME. The PLMN ID is defined in the attribute enodebplmnid in the parent ENodeBFunction. PLMN IDs from additionalplmnlist also appear in this list. The list may include up to 6 PLMNs. administrativestate = LOCKED The administrative state. advcellsupaction = NO_ACTION Controls which recovery actions to be performed when ACS detects a sleeping cell. LZT R1A Ericsson AB

96 LTE L13 Radio Network Functionality advcellsupsensitivity = 0 { } Indicates the sensitivity of the Advanced Cell Supervision function. A higher value will make ACS detect a sleeping cell faster, but with higher risk for false detection. A lower value will make ACS detect a sleeping cell slower, but with lower risk for false detection. cellbarred = NOT_BARRED Indicates if the cell is barred and should not be accessible to random UEs. 4.1 Cell Barring and Cell Reservation Cell barring and cell reservation for the operator's use are two conditions that can be configured to exclude a cell as a candidate for cell reselection. A UE is not allowed to select or reselect a cell configured is barred. The UE is allowed to select or reselect another cell on the same frequency, if the cell selection or reselection criteria for the other cell are fulfilled. The UE excludes the cell from the candidate list for five minutes (300 seconds) after it has determined the cell is barred. Similarly, if a cell is configured as reserved for operator use, a UE considers the cell is barred, unless the UE belongs to access classes 11 or 15 (see definition below) and operates in its home PLMN or an equivalent home PLMN. A UE belonging to the access classes 11 or 15 and operating in the home PLMN or equivalent home PLMN considers the cell configured as reserved for operator use as any other cell and may treat the cell as a candidate for cell reselection. Limited services, which are obtained when the UE is camping on an acceptable cell. Important Related Attributes: acbarringforemergency = false Access class barring for AC 10 acbarringformodata Access class barring parameters for mobile originating calls. The information in broadcasted in SIB Ericsson AB 2012 LZT R1A

97 Radio Connection Supervision acbarringformodatapresent = false Specifies presence of Information Element ac-barringformo-data in SIB2. acbarringformosignalling Access class barring parameters for mobile originating signalling. The information in broadcasted in SIB2. acbarringformosignallingpresent = false Specifies presence of Information Element ac-barringformo-signalling in SIB2. acbarringinfopresent = false Specifies presence of Information Element ac-barringinfo in SIB2. Operator-related services, which allow the operator to test newly deployed cells without being disturbed by normal traffic. The operator can establish cell access restrictions by the cellreservedforoperatoruse cell parameter that allows the reservation of a cell for operator use only. In order to restrict the number of simultaneous access attempts, it is also possible to restrict access for certain access classes and access types using acbarringinfo. Access class barring offers a possibility to deny access for a certain fraction of UE to save network resources, for example, in situations where network resources have the risk to become congested. Access class barring allows normal service to be maintained for UE of special access classes and for the fraction of UE where access is granted Definition of Access Classes The access class to which the UE belongs is stored on the Universal Subscriber Identity Module (USIM). It is tied to the subscription. All UE (with an USIM) are members of one of ten randomly allocated populations, defined as access classes. LZT R1A Ericsson AB

98 LTE L13 Radio Network Functionality SIB2 : Access Class Barring Emergency Call Signalling Data Ordinary Categories: Class 0-9 Special Categories: Class 15: PLMN staff Class 14: Emergency Services Class 13: Public Utilities Class 12: Security Services Class 11: For PLMN Use Figure 3-12 Access Classes In addition, a UE may be member of one or more of five special categories (access classes 11-15). Those classes are allocated to specific user categories according to 3GPP TS , as follows: Class 15: PLMN staff Class 14: Emergency Services Class 13: Public Utilities Class 12: Security Services Class 11: For PLMN Use Access classes 0-9 apply in any home or visited PLMN. Access classes 11 and 15 apply in the home PLMN and any equivalent home PLMN, Figure Access classes apply in any home PLMN or visited PLMN in the home country. The home country is defined by the Mobile Country Code (MCC) part of the International Mobile Subscriber Identity (IMSI) stored on the USIM. There is no access class 10 in a true sense. (AC 10 could be considered as the implicit AC the UE assumes during an emergency call Access Class Restriction Information The right to perform certain types of access in the cell is controlled by the acbarringinfo broadcast in system information (SIB type 2). It applies to mobile originating access. The acbarringinfo information distinguishes mobile originating access for the purpose of initiating the following: Ericsson AB 2012 LZT R1A

99 Radio Connection Supervision An emergency call (not supported in Rel-8, but the barring information is included in sys info, as a preparation for Rel-9) A connection for signaling purposes A connection for user data transfer Emergency Call Barring Emergency call barring is a condition that, when configured, prevents the UE from initiating an emergency call in the cell. The emergency call barring is configured, if the access class restriction information is present in the system information and the emergency call barring in the cell is set to the value TRUE. When emergency calls are barred, UEs are not allowed to establish emergency calls in the cell. The only exceptions are UEs belonging to a special access class 11 to 15. If the access class restriction information for mobile originating data is present in the system information and the sending of mobile originating data is not restricted for the special access class the UE belongs to, these UEs may establish emergency calls. When emergency calls are allowed in the cell, emergency calls are allowed for any UE, including UE without a USIM (that is, UE without an IMSI) and UE in camped on any cell state (that is, UE unable to register in the PLMN or in the tracking area to which the cell belongs) Access Class Restriction of Signaling and User Data The access class restrictions are applied separately for signaling and mobile originating user data. If the applicable access class restriction information is not present, access class restrictions do not apply. When access class restrictions are applied for the special access classes 11-15, the network indicates if access to the cell is restricted or not for each access class. If the UE belongs to one of the special access classes and the access to the cell for that access class is indicated as not restricted, the UE may establish the RRC connection and does not consider the access class restrictions for access classes 0-9. When access class restrictions are applied for the access classes 0-9, the network indicates a probability factor, which applies to all the access classes 0 to 9. The UE performs a random draw at the attempt to establish an RRC connection. If the result is less than the indicated probability factor, access to the cell is granted and the UE may establish the RRC connection. Otherwise, access to the cell is restricted, preventing the UE from establishing the RRC connection. LZT R1A Ericsson AB

100 LTE L13 Radio Network Functionality When access class restrictions deny access to the cell, the UE considers access to the cell as restricted for further access during a randomized back-off time. The average back-off time is indicated in system information. The UE applies an actual back-off time, which is randomly distributed with an equal probability within +/- 30 percent of the indicated average. If the UE reselects to another cell during the back-off time, the access class restrictions and the back-off time from the previous cell are cleared and the UE may attempt access in the new cell. The access class restrictions do not apply to mobile terminating user data. However, if access to the cell is prohibited due to a previous attempt to send signaling or mobile originating user data, the restriction during the back-off time applies also to mobile terminating user data. 5 Cell Selection and reselection The cell selection and reselection process allows the UE to look for a suitable cell in the selected PLMN and to camp on it. When a suitable cell is found, the UE camps on it in a state defined as camped normally. In this state, the UE monitors paging and system information, performs periodical radio measurements, and evaluates cell reselection criteria. If the UE finds a better cell, that cell is selected in the cell reselection process. The change of cell may imply a change of the RAT. There are two different cases - initial cell selection and stored information cell selection, as shown in Figure no suitable Cell found Cell Selection when leaving connected mode return to Idle Mode Connected Mode cell information stored for the PLMN Stored Information Cell Selection suitable cell found suitable cell found go here whenever a new PLMN is selected 1 no cell information stored for the PLMN Initial no suitable cell found Cell Selection 2 Camped suitable cell found Normally TS Leave Idle Mode Selected PLMN Suitable trigger is rejected Cell found go here When no USIM Cell Reselection Evaluation Process no suitable Cell found Any Cell Selection in the UE no acceptable cell found USIM inserted Cell Selection when leaving connected mode return to Idle Mode Connected Mode (Emergency calls only) Acceptable Cell found leave Idle Mode Camped on any cell Acceptable trigger Cell found Cell Reselection Evaluation Process Acceptable Cell Found Suitable Cell found no acceptable Cell Found Ericsson AB 2012 LZT R1A

101 Radio Connection Supervision Figure 3-13 The cell selection process. The UE uses one of the following two cell selection procedures: a) Initial Cell Selection This procedure requires no prior knowledge of which RF channels are E-UTRA carriers. The UE shall scan all RF channels in the E-UTRA bands according to its capabilities to find a suitable cell. On each carrier frequency, the UE need only search for the strongest cell. Once a suitable cell is found this cell is selected. b) Stored Information Cell Selection This procedure requires stored information of carrier frequencies and optionally also information on cell parameters, from previously received measurement control information elements or from previously detected cells. Once the UE has found a suitable cell the UE shall select it. If no suitable cell is found the Initial Cell Selection procedure is started. The Cell Selection process is run when: When the UE is switched on When the UE returns to Idle mode from connected mode After a number of failed attempts of RRC connection request when the UE is in idle mode and tries to establish a RRC Connection Cell Selection procedures: Initial Cell Selection scan all RF channels Stored Information Cell Selection use stored info on carrier frequencies and cell parameters Figure 3-14 Cell Selection Cases For PLMN selection and to find a suitable cell, it is necessary for the UE to synchronize to the E-UTRAN to read the system information on the BCCH. The synchronization between the UE and E-UTRAN is obtained by the cell search procedure. LZT R1A Ericsson AB

102 LTE L13 Radio Network Functionality 5.1 Cell search procedure The cell search procedure allows the UE to acquire the slot and frame synchronization and to get the downlink cell ID and cell ID group associated with the cell. The Physical Cell ID (PCI) can then be calculated by the UE. The physical signals involved in the cell search procedure are the Primary and Secondary Synchronization Signal (P-SS and S-SS). The procedure is based on the steps shown in Figure Detection of carrier frequency P-SS Detection of SS symbol timing Identification of cell ID (N ID,2 =0,1,2) Detection of radio frame timing Cell Search S-SS Detection of cell id group (N ID,1 =0,...,167)=> PCI Detection of MIMO & CP configuration Possible to read Sys Info & RS (timing, seq, freq shift) Note: PCI= 3 N ID,1 +N ID,2 Figure 3-15: Cell Search Procedure Ericsson AB 2012 LZT R1A

103 Radio Connection Supervision 5.2 Cell selection procedure When the UE has information on the carrier frequencies of the PLMN previously stored on the USIM, it may use the stored information cell selection procedure. This information ensures the simplification of the cell search procedure and speeds up the search for a suitable cell. After synchronization, the UE reads system information on the BCCH. If all requirements for a suitable cell are fulfilled, the UE selects that cell and tries to register. If the UE does not find a suitable cell in the PLMN on which it was previously registered, and it has already used the stored information cell selection procedure, the initial cell selection procedure is started. This procedure does not require knowledge of radio frequency channels E-UTRAN band. The UE scans all radio frequency channels to find a suitable cell. On each carrier, the UE searches for the cell with the highest signal level, according to the cell search procedure, and reads the system information on the BCCH Cell Selection Criterion The cell selection criterion S is fulfilled when: Srxlev > 0 Where: Srxlev = Q rxlevmeas (Q rxlevmin + Q rxlevminoffset ) - Pcompensation Where: the signaled value QrxlevminOffset is only applied when a cell is evaluated for cell selection as a result of a periodic search for a higher priority PLMN while camped normally. During this periodic search for higher priority PLMN the UE may check the S criteria of a cell using parameter values stored from a different cell of this higher priority PLMN. LZT R1A Ericsson AB

104 LTE L13 Radio Network Functionality Table 1 Cell Selection Parameters as defined in Srxlev Q rxlevmeas Q rxlevmin Q rxlevminoffset Pcompensation P EMAX P UMAX Cell Selection RX level value (db) Measured cell RX level value (RSRP). Minimum required RX level in the cell (dbm) Offset to the signalled Q rxlevmin taken into account in the Srxlev evaluation as a result of a periodic search for a higher priority PLMN while camped normally in a VPLMN max(p EMAX P UMAX, 0) (db) Maximum TX power level an UE may use when transmitting on the uplink in the cell (dbm) defined as P EMAX in [TS ] Maximum RF output power of the UE (dbm) according to the UE power class as defined in [TS ] The cell selection criterion (the S criterion) is based on the measured Reference Signal Received Power (RSRP) level in the cell, normalized with respect to the minimum required receive signal level in the cell and certain compensations. For further information, see 3GPP TS The parameters used to control the cell selection process are: qrxlevmin which corresponds to Qrxlevmin and pmaxservingcell which corresponds to P EMAX When camped on a cell, the UE regularly searches for a better cell according to the cell reselection criteria. If a better cell is found, that cell is selected. The change of cell may imply a change of RAT. For normal service, the UE camps on a suitable cell, tunes to that cell's control channel(s) so that the UE can: Receive system information from the PLMN; and receive registration area information from the PLMN, e.g., tracking area information; and receive other AS and NAS Information; and if registered: receive paging and notification messages from the PLMN; and initiate transfer to connected mode Ericsson AB 2012 LZT R1A

105 Radio Connection Supervision One of the requirements for a suitable cell is that it fulfills the cell selection criterion. The UE bases its evaluation on RSRP (Reference Signal Received Power). The cell selection criterion is fulfilled when: S-criterion Srxlev = Qrxlevmeas qrxlevmin + QRxLevMinOffset Pcompensation* > 0 measured RSRP Parameter (-140 to -44) Parameter (0 to 16) *Pcompensation = max(pmaxservingcell P;0) Parameter (-30 to 33 or 1000) UE Power: 23 dbm S>0 S>0 S>0 S>0 S>0 S>0 S>0 S>0 Figure 3-16 Cell Selection (S-Criterion) S>0 S>0 S>0 S>0 S>0 S>0 S>0 qrxlevmin is sent in the system information (SIB 3 for intra frequency cells) and indicates the minimum required signal strength. The UE measures the Reference Signal Received Power (RSRP and obtains Srxlev. qrxlevminoffset is an offset to Qrxlevmin taken into account at periodic search for a higher priority PLMN. pmaxservingcell (serving cell) and pmax (neighboring cell); if absent/not sent in SIB1 (serving) or SIB3 (inter frequency neighbor), the UE applies the maximum power according to its capability. P = P UMAX is the UE maximum output power according to its class. LZT R1A Ericsson AB

106 LTE L13 Radio Network Functionality Quantities measured and parameters used are shown below. qrxlevmin pmaxservingcell/pmax Qrxlevmeas Reference Signal BCCH Figure 3-17: Determination of Cell Selection Criterion The UE measures the RSRP of the serving cell and evaluates the cell selection criterion, S, at least every DRX cycle, see Figure The UE measures the RSRP of the serving cell at least every DRX Cycle UE evaluate N serv consecutive DRX Cycles for cell reselection DRX cycle length [s] N serv [number of DRX cycles] Figure 3-18 Measurement Intervals. The UE shall filter the RSRP measurements of the serving cell using at least 2 measurements. Within the set of measurements used for the filtering, at least two measurements shall be spaced by, at least DRX cycle/ Ericsson AB 2012 LZT R1A

107 Radio Connection Supervision If the UE has evaluated in N serv consecutive DRX cycles that the serving cell does not fulfill the cell selection criterion S, the UE shall initiate the measurements of all neighbor cells indicated by the serving cell, regardless of the measurement rules currently limiting UE measurement activities. Filtered Qrxlevmeas...N serv Qrxlevmeas < 0 Qrxlevmeas<0 Qrxlevmeas<0 Qrxlevmeas<0 Qrxlevmeas Qrxlevmeas (neighbor) Qrxlevmeas (neighbor) (neighbor) OUT OF SERVICE... drxcycle drxcycle drxcycle drxcycle 10s time Cell Reselection Initial Cell Selection Figure 3-19 Measurement Intervals, Example If the UE in RRC_IDLE has not found any new suitable cell based on searches and measurements using the intra-frequency, inter-frequency and inter-rat information indicated in the system information for 10 s, the UE shall initiate cell selection procedures for the selected PLMN. After this 10 s period a UE in RRC_IDLE state is considered to be "out of service area" and shall perform actions according to this. If both S- and R-criteria (cell reselection criterion) are fulfilled and other requirements for a suitable cell are fulfilled, the UE will camp on the cell. The UE enters the state camped normally where it performs intra-, and inter-system radio measurements to evaluate if a neighboring cell is better than the serving one. The UE in idle mode measures the RSRP level of the E-UTRA cells on the current frequency and any inter-frequency carriers and the signal levels of cells on any inter-rat frequency. These measurements are performed in the camped normally and camped on any cell states, depending on parameters broadcast in system information. The results from these measurements are used in the cell reselection evaluation process. The amount of measurements the UE performs may impact UE battery performance. A trade-off between cell reselection and UE battery performance is required when setting configuration parameter values to control these measurements. The criteria for performing the different kinds of measurements are based on the SServingCell value, which is the same as the Srxlev value of the serving cell. LZT R1A Ericsson AB

108 LTE L13 Radio Network Functionality Measurements on intra-frequency cells are performed if SServingCell SIntraSearch. If SServingCell > SIntraSearch, the UE does not need to perform these measurements. Measurements on E-UTRA cells on inter-frequency carriers and cells on inter- RAT frequencies with higher cell reselection priority than the current frequency are always performed. Measurements on E-UTRA cells on inter-frequency carriers and cells on inter- RAT frequencies with equal or lower cell reselection priority than the current frequency are performed if SServingCell SnonIntraSearch. If SServingCell > SnonIntraSearch, the UE does not need to perform these measurements. 5.3 Cell reselection procedure The cell reselection procedure is valid for UEs in idle mode and intra-frequency scenarios. In order to always camp on the best cell the UE performs the cell reselection procedure in the following cases: The Cell Reselection process is run when: When the cell on which it is camping is no longer suitable. When the UE, in camped normally state, has found a better neighboring cell than the cell on which it is camping. When the UE is in limited service state on an acceptable cell. S ServingCell can be used for disabling mesurements on intra- (SIntraSearch) or inter- (SNonIntraSearch) frequency neighbors: S ServingCell = Srxlev S ServingCell <= SIntraSearch: perform measurements S ServingCell > SIntraSearch: no measurements Figure 3-20 Cell Reselection Cases When the UE triggers a cell reselection evaluation process, it ranks cells that fulfill the cell selection criteria. The UE ranks the cells according to the R- criteria. The UE performs ranking of all cells fulfilling the S-criterion (S > 0). The UE ranks the cells according to the R-criterion as illustrated in Figure 2-20 below Ericsson AB 2012 LZT R1A

109 Radio Connection Supervision R-criteria 0 to 24 db R(serving) = Qmeas(s) + + Qoffmbms* R(serving) = Qmeas(s) + qhyst R(neighbor) = Qmeas(n) - Qoffset R(neighbor) Rank 1 R(neighbor) Rank 2 R(neighbor) Rank 3 Where Qoffset is: qoffsetcelleutran: Cell individual offset in the intrafrequency and equal priority inter-frequency cell ranking criteria qoffsetfreq: Frequency specific offset in the equal priority inter-frequency cell ranking criteria treselectioneutra Figure 3-21 Cell Reselection (R-Criteria) Qmeas is the quality value of the received signal. (RSRP) qhyst(s) is the hysteresis value (qhyst) that is read in the system information of the serving cell. Qoffset is an offset in the cell ranking criterion of neighbor E- UTRA cells. It consists of a cell individual part and a frequency specific part. The frequency specific part applies to equal priority inter-frequency cells only. The UE reselects the new cell, if the cell reselection criteria are fulfilled during the time interval treselectioneutra. LZT R1A Ericsson AB

110 LTE L13 Radio Network Functionality 5.4 Cell Reselection evaluation process The UE only performs cell reselection evaluation for E-UTRAN frequencies and inter-rat frequencies that are given in system information and for which the UE has a priority provided. The UE does not consider any black listed cells as candidate for cell reselection. RSRP qhyst(s) sintrasearch snonintrasearch Qmeas(n) R(n) qoffset(s) R(s) Qmeas(s) treselectioneutra time Cell reselection Figure 3-22 Cell Reselection Evaluation Process Measurement rules for Intra-frequency and equal priority inter-frequency cell re-selection The amount of measurements the UE performs may impact UE battery performance. A trade-off between cell reselection and UE battery performance is required when setting configuration parameter values to control these measurements. The criteria for performing the different kinds of measurements are based on the S criterion of the serving cell. S ServingCell = Qrxlevmeas - Qrxlevmin Measurements on intra-frequency cells are performed if S ServingCell <= Sintrasearch If S ServingCell > SIntrasearch, the UE does not need to perform these measurements Ericsson AB 2012 LZT R1A

111 Radio Connection Supervision When camped on an E-UTRA cell, the UE performs a ranking of neighboring E- UTRA cells on the same frequency. The cell ranking is based on the Reference Signal Received Power (RSRP) measurement quantities of the serving cell (Qmeas,s) and the neighboring cells (Qmeas,n) satisfying the cell selection criteria. The UE applies the cell ranking criterion Rs on the serving cell, and the cell ranking criterion Rn on the intra-frequency cells: R s = Q meas, s + Q Hyst R n = Q meas, n - Qoffset where: Q meas Qoffset RSRP measurement quantity used in cell reselections. For intra-frequency: Equals to Qoffset s,n, if Qoffset s,n is valid, otherwise this equals to zero. For inter-frequency: Equals to Qoffset s,n plus Qoffset frequency, if Qoffset s,n is valid, otherwise this equals to Qoffset frequency. The Qhyst is a hysteresis value preventing too frequent reselection back and forth between cells of nearly equal rank. When a neighboring cell is ranked as better than the serving cell (that is, Rn > Rs) during a time interval treselectioneutra, the UE performs a cell reselection to the better ranked cell. UE measures on serving cell and neighboring cells (IRAT & intra-lte) according to list and considers cells that fulfill cell reselection criterion RS: UE measures RSRP SI: cell reselection parameters RS: UE measures RSRP Serving cell UE performs cell reselection autonomously based on measurements Cell reselection? treselectioneutra Rn > Rs? R s = Q meass + Q Hyst R n = Q measn - Q offset Neighboring cell Figure Idle Mode Mobility. LZT R1A Ericsson AB

112 LTE L13 Radio Network Functionality 5.5 Priority based cell reselection When multiple E-UTRA and/or inter-rat frequencies are used, priority based cell reselection can be applied. Information about the available E-UTRA and inter-rat frequencies, the cell reselection priorities and the threshold values for cell reselection is provided to the UE via the system information in the cell where the UE is camping. The UE behavior depends on the cell reselection priorities of those frequencies. The UE only perform cell reselection evaluation for the E-UTRA and inter-rat frequencies that are given in system information and for which a cell reselection priority is provided. If the UE finds an inter-frequency carrier or an inter-rat frequency with a cell reselection priority higher than the frequency where the UE is camping, the UE attempts to reselects a cell on that frequency. Cell reselection occurs if the UE finds a cell with a Srxlev value greater than the threshxhigh value for that frequency. If the UE finds an inter-frequency carrier with equal priority to the frequency where the UE is camping, the UE performs cell reselection much in the same way as intra-frequency cell reselection. A UE camping in E-UTRA is not required to evaluate an equal priority inter- RAT frequency. Configuration of equal cell reselection priority of inter-rat frequencies should be avoided. The UE behavior is undefined.if the Srxlev value of the serving cell falls below the threshservinglow value, the UE attempts to reselect a cell on an inter-frequency carrier or an inter-rat frequency with cell reselection priority lower than the frequency where the UE is camping. Cell reselection occurs if the UE finds a cell with an Srxlev value greater than the threshxlow value for that frequency. Neighbor cellreselectionpriority higher than for used freq Frequency 1 R(serving) and = Qmeas(s) + + Qoffmbms* Prio 1 Srxlev > threshxhigh triggers cell reselection to higher prio frequency (E-UTRAN or IRAT) Frequency 2 Prio 2 Srxlev < threshservinglow and Frequency 3 Srxlev > threshxlow triggers cell reselection to lower Prio 3 prio frequency (E-UTRAN or IRAT) Configuration of equal priority IRAT frequencies should be avoided! Figure 3-24 Priority Based Cell Reselection Ericsson AB 2012 LZT R1A

113 Radio Connection Supervision 5.6 Speed-Dependent Scaling of Cell Reselection The speed-dependent scaling of cell reselection criteria is used to influence the cell reselection criteria for fast moving UE. It helps the UE to respond more quickly to cell changes when moving at high speed. A UE may enter three different mobility states: Normal Medium High mobility The usual treselectioneutra and qhyst parameters are used in the normal mobility state for the evaluation of cell reselection criteria. In the medium and high mobility states, the UE applies a scaling factor, decreasing the value of treselectioneutra parameter (treselectioneutrasfmedium and treselectioneutrasfhigh). In that way, the evaluation period of cell reselection criteria is reduced. In addition, a negative offset is added to the qhyst hysteresis value (qhystsfmedium and qhystsfhigh) in the cell ranking criteria. It lowers the threshold for the reselection of intra-frequency cells. The criteria for the UE to enter the medium and high mobility states are based on the number of recent cell reselections performed by the UE. A sliding time window is used. The parameter tevaluation determines the duration of the sliding time window. The parameters ncellchangemedium (medium mobility) and ncellchangehigh (high mobility) determine the number of cell reselections the UE performs within the sliding time window. The UE applies an additional time period before reentering the normal mobility state LZT R1A Ericsson AB

114 LTE L13 Radio Network Functionality The diagram below illustrates the speed dependent scaling of Cell Reselection. Three mobility states: Normal, Medium and High Based on the no of cell reselections made by the UE ncellchangemedium ncellchangehigh tevaluation RSRP Normal: - treselectioneutra - qhyst RSRP Medium: - treselectioneutramedium - qhystmedium High: - treselectioneutrahigh - qhysthigh sintrasearch sintrasearch qhyst Qmeas(n) qhystsf Medium / qhystsf High Qmeas(n) R(n) R(n) qoffset(s) R(s) qoffset(s) R(s) Qmeas(s) Qmeas(s) treselectioneutra time Cell reselection time treselection EutraSfMedium / treselectioneutra SfHigh Cell reselection Figure 3-25 Speed Dependent Scaling of Cell Reselection Once in the medium or high mobility state, the UE re-enters the normal mobility state when the number of cell reselections during the sliding time window (tevaluation) stays below the parameter ncellchangemedium and ncellchangehigh values during a period equal to the additional time period. The parameter thystnormal determines the duration of the additional time period. N CR /tevaluation ncellchange High ncellchange Medium thystnormal Enter Medium Enter High Enter Medium Enter Normal time Figure 3-26 Speed Dependent Scaling of Cell Reselection Ericsson AB 2012 LZT R1A

115 Radio Connection Supervision 5.7 Tracking Area Update A TA is the set of E-UTRA cells in a PLMN, identified by a common Tracking Area Code (TAC) in the system information. When a UE registers itself in the network, the core network stores information about the tracking area where the registration is performed. This information is used, for example, to assist the UE paging. The tracking area update procedure is used by the UE to update the registration of its actual location in the network. The core network provides the UE with a list of tracking areas where the registration is valid. The UE performs a new registration, either after a certain time (periodic registration), or when it enters a new tracking area where the registration is no longer valid. The operator configures the TAC associated with each cell. 5.8 Access Class Restrictions Access class restrictions are conditions that can be configured to prevent the UE from performing certain types of access in the cell. These conditions are based on the The access class to which the UE belongs The type of access to be perform Access class restrictions take effect when the UE leaves idle mode and performs the RRC connection establishment procedure to enter connected mode. If access is prevented due to access class restrictions, the UE considers the cell to be temporarily barred for initial access and it aborts the intended RRC connection establishment procedure. The UE remains in idle mode. 5.9 Paging Paging is used by the network to communicate with User Equipment (UE) in idle mode. In these situations, the network does not know on which cell the UE is camped. The LTE network uses Paging to notify UE in idle mode of an incoming data session, system information change and ETWS notifications. LZT R1A Ericsson AB

116 LTE L13 Radio Network Functionality CN Initiated Paging enb Initiated Paging MME S1AP Paging message TAC 2 TAC 1 RRC Paging message Figure 3-27 Paging In case of CN initiated Paging the enb receives the S1AP Paging message from the MME and determines the paging occasion where the UE monitors PCCH. The UE paging identities are queued separately for each PO. In case of enb initiated paging due to ETWS notification or System Info change Paging is sent in all paging occasions E-UTRAN initiates the paging procedure by transmitting the Paging message at the UE s paging occasion. In case of a paging overload, the enb discards paging identities that have not been transmitted in due time. The Paging message may address up to 16 UEs in the same PO, but the operator may restrict this number using the maxnoofpagingrecord configuration parameter. The UE is normally addressed using the S-TMSI. The UE may use Discontinuous Reception (DRX) in idle mode in order to reduce power consumption Ericsson AB 2012 LZT R1A

117 Radio Connection Supervision The UE may use Discontinuous Reception (DRX) in idle mode in order to reduce power consumption and subsequently increase battery life. UE reads SIB2 (IE radioresourceconfigcommon) to calculate when to wake to monitor the Paging channel. When DRX is used the UE needs only to monitor one P-RNTI per DRX cycle. UE monitors PDCCH channel at the start of the subframe to see if a paging message is included PF PO If the UE detects a P-RNTI value coded with the CRC of the PDCCH then it will decode the PDSCH to see if the paging message is intended for it PF UE receiver circuitry switched off DRX cycle Figure 3-28 Paging DRX Cycle One Paging Occasion (PO) is a subframe where there may be P-RNTI transmitted on PDCCH addressing the paging message. One Paging Frame (PF) is one Radio Frame, which may contain one or multiple Paging Occasion(s). When DRX is used the UE needs only to monitor one PO per DRX cycle. PF and PO is determined by following formula using the DRX parameters provided in System Information. PF is given by following equation: SFN mod T= (T div N)*(UE_ID mod N) Index i_s pointing to PO from subframe pattern defined in the table below will be derived from following calculation: LZT R1A Ericsson AB

118 LTE L13 Radio Network Functionality i_s = floor(ue_id/n) mod Ns PO when i_s=0 PO when i_s=1 PO when i_s=2 PO when i_s=3 1 9 N/A N/A N/A N/A N/A System Information DRX parameters stored in the UE shall be updated locally in the UE whenever the DRX parameter values are changed in SI. If the UE has no IMSI, for instance when making an emergency call without USIM, the UE shall use as default identity UE_ID = 0 in the PF and i_s formulas above. The following Parameters are used for the calculation of the PF and i_s: T: DRX cycle of the UE. T is determined by the shortest of the UE specific DRX value, if allocated by upper layers, and a default DRX value broadcast in system information. If UE specific DRX is not configured by upper layers, the default value is applied. nb: 4T, 2T, T, T/2, T/4, T/8, T/16, T/32. N: min(t,nb) Ns: max(1,nb/t) UE_ID: IMSI mod IMSI is given as sequence of digits of type Integer (0..9), IMSI shall in the formulae above be interpreted as a decimal integer number, where the first digit given in the sequence represents the highest order digit. Parameter nb is used to calculate the number and position of Paging Occasions (PO) and Paging Frames (PF). The numerical value of nb is dependent on the value of the defaultpagingcycle (T). When nb is set to T, 2T or 4T, it determines the number of POs per PF, and the PO position in PF. When nb is set to a value smaller than T, it affects the System Frame Number of the PF, position of PO in the PF, and also distribution of User Equipments into groups with the same PF. When is set to a smaller value, the groups will be fewer but larger and vice versa Ericsson AB 2012 LZT R1A

119 Radio Connection Supervision PF index when: SFN mod T= (T div N)*(UE_ID mod N) i_s = floor(ue_id/n) mod Ns defaultpagingcycle (T): 32, 64, 128, 256 radio frames. nb: 4T, 2T, T, 1/2T, 1/4T, 1/8T, 1/16T, 1/32T N: min(t,nb) Ns: max(1,nb/t) UE_ID: IMSI mod ms PF T=pagingDefault [32,64,128,256] PF Example: T = 64 nb = 2T = 128 UE_ID = IMSI mod(1024) PF SFN 0 SFN 64 SFN 960 N = min(t,nb) = 64 Ns = max(1,nb/t) = 2 UE_ID = IMSI mod 1024 = e.g 0 PF : SFN mod T= (T div N)*(UE_ID mod N) = 0, 64, 128 1ms Subframe number Ns i_s=0 i_s=1 i_s=2 i_s=3 Subframe 0 Subframe 4 Subframe N/A 9 N/A N/A N/A N/A PDCCH PDSCH Figure 3-29 DRX Paging Frame The paging frame and paging occasion structure is shown in an example below. defaultpagingcycle=rf32 Paging Frame nb = 4T nb = 2T nb = T nb = (½)T nb = (¼)T nb = (1/8)T nb = (1/16)T nb = (1/32)T => => => => => => => => Paging Occasion (per Paging Frame) Figure 3-30: Paging Frame and Paging Occasion example LZT R1A Ericsson AB

120 LTE L13 Radio Network Functionality Adaptive paging is supported since L11. Adaptive paging is described in the following picture: Allows the operator to configure if a first page is to be distributed to the area of a single RBS, TA or the whole TA list The selected area in which the page is sent can be decided based on different criteria: UE related criteria, IMEI series and IMSI series. These criteria are used to send a page to a single RBS or a TA for known stationary UEs such as electricity meters Time since UE last reported location. This criterion is used to send a page to a single RBS or a TA when a the location of the UE is unknown Service related criteria, APN, QCI and ARP. These criteria are used for time critical services to guarantee that the page is sent directly to all RBSs in TA list The criteria are assigned a priority value which indicates the order the criteria are verified. Figure 3-31: Adaptive Paging 6 Location registration The TA (-list) is an area in which the CN sends a paging message. One TA (-list) is a set of cells, configured by the operator. Each cell belongs to one and only one TA (in each PLMN, if shared network), configured by the operator. At registration, the UE receives a list of TA identities (TAI list) defining the area where the registration is valid. Paging need to cover all TAs in the TAI list. The TAI (PLMN + TAC) of the TA the cell belongs to is signaled in system information. Three different types of registration updates: Normal registration Periodic Registration Attach/detach Figure 3-32 Location Registration Ericsson AB 2012 LZT R1A

121 Radio Connection Supervision 6.1 Normal registration A Normal Registration update is done when: The UE is switched on The UE is moving into a new TA (-list) Valid in both connected and idle mode Figure 3-33 Normal Registration The UE reads TAI in system information and detects that the received area identities differ from the ones stored on the USIM. If the TAI received from the system information is not in the forbidden TAIs list, a TA update request is sent by the UE to the CN. If the TAI is forbidden, the UE tries to select another cell belonging to a permitted TAI or another PLMN. The number of TA update requests is controlled by an attempt counter, in order to limit the number of location update attempts made by UE, when location updating is unsuccessful. After successful location update, the UE stores the list of TAIs in the USIM and resets the attempt counter. 6.2 Periodic Registration Periodic registration (TA updating) is used to locate the UE to avoid unnecessary paging attempts for a UE that has lost coverage and is not able to inform the CN that it is inactive. A timer t3412 controls the periodic TA update procedure and gives the time interval between two consecutive periodic location area updates. The timer is sent in the TAU accept message from the MME. T3412 is stopped when entering EMM-CONNECTED mode and reset and started when going from EMM-CONENCTED mode to EMM-IDLE mode. See Figure UE in idle mode UE moves to connected mode UE moves to idle mode t3412 t3412 t3412 TA Update TA Update TA Update Figure 3-34 Periodic Registration LZT R1A Ericsson AB

122 LTE L13 Radio Network Functionality 7 PARAMETERS 7.1 Attributes of MO EUtranCellFDD to configure Idle Mode support Parameter cellbarred Description Parameter specifying whether the cell is barred and not accessible by UEs (cellbarred in SIB type 1). cellreservedforoperatoruse Range: Barred, Not barred Parameter specifying whether the cell is reserved for operator use, for UE belonging to access classes 11 or 15 (cellreservedforoperatoruse in SIB type 1). Not reserved qrxlevmin Required minimum RSRP level in the E UTRA cell (q-rxlevmin in SIB type 1). Corresponds to parameter Q rxlevmin in the document 3GPP TS Unit: dbm Range: -140 to -44 qrxlevminoffset pmaxservingcell Resolution: 2 Offset to qrxlevmin taken into account at periodic search for a higher priority PLMN. Maximum UE power use in the serving cell (p-max in SIB type 1). If absent, the UE applies the maximum UE power for the UE power class. Corresponds to parameter P EMAX in the document 3GPP TS Unit: dbm SIB3 ncellchangehigh Range: -30 to 33 MO Attributes in Structure SIB3 Number of cell changes to enter the high mobility state (n-cellchangemedium in SIB type 3). Corresponds to parameter N CR_H in the document 3GPP TS ncellchangemedium Range: 1 16 Number of cell changes to enter the medium mobility state (n-cellchangehigh in SIB type 3). Corresponds to parameter N CR_M in the document 3GPP TS qhyst Range: 1 16 Cell reselection parameter that defines the hysteresis value in the intra-frequency cell ranking criteria (q-hyst in SIB type 3). Corresponds to parameter Q hyst in the document 3GPP TS Unit: db qhystsfhigh Range: 0, 1, 2, 3, 4, 5, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24 Reduction of the Qhyst parameter applied in high mobility state (q-hystsf.sf- High in SIB type 3). Corresponds to parameter sf-high of Speed dependent ScalingFactor for Q hyst in the document 3GPP TS Unit: db Range: -6 to 0 Resolution: Ericsson AB 2012 LZT R1A

123 Radio Connection Supervision qhystsfmedium Reduction of the Q hyst parameter applied in medium mobility state (q-hystsf.sf- Medium in SIB type 3). Corresponds to parameter sf-medium of Speed dependent ScalingFactor for Q hyst in the document 3GPP TS Unit: db Range: -6 to 0 Resolution: 2 sintrasearch Threshold for intra-frequency measurements (s-intrasearch in SIB type 3). Corresponds to parameter S intrasearch in the document 3GPP TS Unit: db Range: 0 62 snonintrasearch Resolution: 2 Threshold for inter-frequency and inter-rat measurements on frequencies of equal or lower priority (s-nonintrasearch in SIB type 3). Corresponds to parameter S nonintrasearch in the document 3GPP TS Unit: db Range: 0 62 tevaluation Resolution: 2 Duration for the evaluation of the entering criteria to the mobility states (t-evaluation in SIB type 3). Corresponds to parameter T CRmax in the document 3GPP TS Unit: s threshservinglow Range: 30, 60, 120, 180, 240 Threshold for inter-frequency and inter-rat measurements on frequencies of lower priority (threshservinglow in SIB type 3). Corresponds to parameter Thresh serving,low in the document 3GPP TS Unit: db Range: 0 62 thystnormal Resolution: 2 Additional duration for the evaluation of the reentering criteria to the normal mobility state (t-hystnormal in SIB type 3). Corresponds to parameter T CRmaxHyst in the document 3GPP TS Unit: s acbarringinfo acbarringfactorformodata Range: 30, 60, 120, 180, 240 MO Attributes in Structure acbarringinfo Probability factor for allowing access classes 0 9 sending mobile originating data (ac- BarringForMO-Data.ac-BarringFactor in SIB type 2). Unit: % Range: 0, 5, 10, 15, 20, 25, 30, 40, 50, 60, 70, 75, 80, 85, 90, 95 Resolution 5 LZT R1A Ericsson AB

124 LTE L13 Radio Network Functionality acbarringfactorformosignalling Probability factor for allowing access classes 0 9 sending mobile originating signaling (ac-barringformo-signalling.ac-barringfactor in SIB type 2). Unit: % Range: 0, 5, 10, 15, 20, 25, 30, 40, 50, 60, 70, 75, 80, 85, 90, 95 Resolution: 5 acbarringforemergency Emergency call barring in the cell (ac-barringforemergency in SIB type 2). acbarringinfopresent Range: boolean Declares the presence of access class restriction information (ac-barringinfo in SIB type 2). acbarringmodatapresent Range: boolean Declares the presence of access class restriction information for mobile originating data (ac-barringformo-data in SIB type 2). acbarringmosignallingpresent Range: boolean Declares the presence of access class restriction information for mobile originating signaling (ac-barringformo-signalling in SIB type 2). acbarringspecialacformodata Range: boolean Access class restrictions for access classes sending mobile originating data (ac- BarringForMO-Data.ac-BarringForSpecialAC in SIB type 2). Range: boolean acbarringspecialacformosignalling Access class restrictions for access classes sending mobile originating signaling (ac-barringformo-signalling.ac-barringforspecialac in SIB type 2). acbarringtimeformodata Range: boolean Average back-off time at access class restriction of mobile originating data (ac- BarringForMO-Data.ac-BarringTime in SIB type 2). Unit: s acbarringtimeformosignalling Range: 4, 8, 16, 32, 64, 128, 256, 512 Average back-off time at access class restriction of mobile originating signaling (ac- BarringForMO-Signalling.ac-BarringTime in SIB type 2). Unit: s Range: 4, 8, 16, 32, 64, 128, 256, Ericsson AB 2012 LZT R1A

125 Radio Connection Supervision 7.2 Attributes of MO EUtranFreqRelation to Configure Idle Mode Support Parameter cellreselectionpriority Description Absolute cell reselection priority for the E-UTRA frequency (cellreselectionpriority in SIB type 3). Corresponds to cell reselection priority in the document 3GPP TS thresxhigh thresxlow thresservinglow pmax qrxlevmin (neighbor cell) qoffsetcelleutran qoffsetfreq treselectioneutrasfhigh treselectioneutrasfmedium Range: 0 7 Threshold for the Srxlev value of the target cell for cell reselection towards a higher priority inter-frequency or IRAT frequency. Threshold for the Srxlev value of the target cell for cell reselection towards a lower priority inter-frequency or IRAT frequency. Threshold for the Srxlev value of the serving cell, below which the UE performs cell reselection towards a lower priority interfrequency or IRAT frequency. Maximum UE power to be used in neighboring cells on the E-UTRA frequency (p-max in SIB type 3). If absent, the UE applies the maximum UE power for the UE power class. Corresponds to parameter P EMAX in the document 3GPP TS Unit: dbm Range: -30 to 33 Required minimum RSRP level in the intra-frequency neighboring cells (q-rxlevmin in SIB type 3). Corresponds to parameter Q rxlevmin in the document 3GPP TS Unit: dbm Range: -140 to -44 Resolution: 2 Cell individual offset in the intra frequency and equal priority inter-frequency cell ranking criteria. Frequency specific offset in the equal priority inter-frequency cell ranking criteria. Scaling factor of the TreselectionEUTRA parameter in high mobility state (t-reselectioneutra-sf.sf-high in SIB type 3). Corresponds to parameter sf-high of Speed dependent ScalingFactor for T reselectioneutra in the document 3GPP TS Range: 0.25, 0.50, 0.75, 1.00 Scaling factor of the T reselectioneutra parameter in medium mobility state (t-reselectioneutra-sf.sf-medium in SIB type 3). Corresponds to parameter sf-medium of Speed dependent ScalingFactor for T reselectioneutra in the document 3GPP TS Range: 0.25, 0.50, 0.75, Paging Parameter maxnoofpagingrecords Description Maximum allowed number of Paging Records included in one RRC paging message, that is, the maximum UEs that can be paged per PO. Default: Refer to Managed Object Model RBS Range: 1-16 Level: RBS LZT R1A Ericsson AB

126 LTE L13 Radio Network Functionality defaultpagingcycle Indicates the number of radio frames in the paging cycle. This is the paging cycle used by the RBS and is broadcast in SIB2. If a UE-specific DRX cycle is provided in the paging message from the MME, that is shorter than the defaultpagingcycle value, then the value from the MME overrides the value in the RBS. The time between POs for user equipment can be calculated by multiplying the defaultpagingcycle parameter value by 10 ms.this parameter corresponds to the variable T shown in the formula in 3GPP TS nb Default: Refer to Managed Object Model RBS Valid values: 32, 64, 128, 256 Level: RBS Used to derive the number of subframes used for paging within each paging cycle. When nb is set to T, 2T or 4T, it affects the number of POs per PFs, and also determines the PO position within PF. When nb is set to 1/2T, 1/4T, 1/8T, 1/16T or 1/32T, it affects the System Frame Number of the PF, the position of the PO within the PF, and distribution of UE into groups with the same PF. When set to a value smaller than T, it affects the System Frame Number of the PF, position of PO in the PF, and distribution of user equipment into groups with the same PF. When nb is set to a smaller value, there are fewer paging groups addressing a larger number of potential user equipment. When nb is set to a larger value, increased paging groups address a decreased number of UEs. pagingdiscardtimer Default: Refer to Managed Object Model RBS Valid values: 4T, 2T, T, 1/2T, 1/4T, 1/8T, 1/16T, 1/32T. Level: RBS Determines the maximum time a received paging message may be retained or queued in the RBS before it is discarded. This timer should be set to the same (or a smaller value) than the Paging-resend timer in the MME (T3413), to prevent the RBS from retaining or sending an old paging message after the resent copy has been received from MME. Default: Refer to Managed Object Model RBS Options: 1-15 Units: s Level: RBS Ericsson AB 2012 LZT R1A

127 Radio Connection Supervision Intentionally Blank LZT R1A Ericsson AB

128 LTE L13 Radio Network Functionality 4 Radio Connection Supervision Objectives After this chapter the participants will be able to: Explain the purpose and function of Radio Connection Supervision 1. Explain how the radio connection supervision is carried out 2. Explain how in-synch and out-of-synch is determined by the radio link monitoring algorithm in the RBS Figure 4-1: Objectives of Chapter Ericsson AB 2012 LZT R1A

129 Radio Connection Supervision 1 Introduction The Radio Connection Supervision (RCS) algorithm supervises the radio connection between E-UTRAN and a UE in connected mode. The purpose of this supervision is to judge whether or not RAN still has control over the UE and to monitor UE s inactivity. Achieve an efficient resource usage in E-UTRAN Avoid any long-term discrepancy concerning the status of the connection between the UE and E-UTRAN Avoid undue charging to customers Figure 4-2. The Main Goals of the Radio Connection Supervision Radio Connection Supervision (RCS) collects radio problems and inactivity information with the purpose to release resources that can be used in a more efficient way. An RRC Connection re-establishment or a UE release may be triggered. In general, it will be determined how serious the condition is and how long it lasts. If Radio Connection Failure (RCF) or Radio Bearer Failure (RBF) is determined, RCS will indicate this to higher layers. The Radio Connection Supervision feature assures that resources in the RBS and between RBS and MME are not reserved for Radio Links or Radio Bearers that are lost or have poor quality and that probably can not be restored. It also monitors inactive UEs. LZT R1A Ericsson AB

130 LTE L13 Radio Network Functionality 2 Overview of radio connection supervision The figure below illustrates the functionalities which RCS depends on. 2. Inactivity Supervision 1. Radio Link Control Radio Connection Supervision Signalling & Payload Bearer Handling Figure 4-3 Overview of Radio Connection Supervision The Signaling and Payload Bearer Handling is responsible for establishment and release of control plane and user plane connections between a UE and the core network. 11 The Radio Link Control (RLC) is responsible for error correction through ARQ, protocol error detection and recovery and more. 12 Inactivity supervision measures the time that a bearer is inactive. This is done in the RLC. The supervision of RC and RB conditions as well as the behavior of UEs can increase the efficiency of the enb and Uu resources by removing RC/RB that probably will not recover and releasing UEs which are inactive for a certain time period. In order to find RCs, RBs and UEs that are probably lost some selected criteria are monitored. In general the criteria are measured in different parts of the enb. They are reported to a central RCS function that indicates the reports as to LTE Signaling and Payload Bearer Handling. The indications can be of the type: Radio Link Failure (RLF) and UE Release due to inactivity Ericsson AB 2012 LZT R1A

131 Radio Connection Supervision Important Impacting Attribute: tinactivitytimer = 61 { 0, } The time a UE can be inactive before it is released. (86400s = 24 hours) Special values: 0 is used to turn off the use of this timer. This means that a UE will not be released due to inactivity. Unit: 1 s Takes effect: New connection. MME PCRF WWW SGW PGW RRC_ CONNECTE D LTE Max 8 SIGNALLING RADIO BEARER S1 BEARER EPS BEARER EPS BEARER EPS BEARER EPS BEARER QCI=5 QCI=1 QCI=9 QCI=2 Inactivity supervision measures the time that a bearer is inactive. This is done in the RLC. Figure 4-4 Inactivity Supervision Important Impacting Attribute: tinactivitytimer = 61 { 0, } The time a UE can be inactive before it is released. (86400s = 24 hours) Special values: 0 is used to turn off the use of this timer. This means that a UE will not be released due to inactivity. Unit: 1 s Takes effect: New connection. 3 Radio Link Monitoring The functionality is called Radio Link Monitoring. The DL supervision of radio conditions will be handled by the UE but some of the parameters that are used for RLF determination by the UE may be set by the enb. The UE will often experience RLF when the enb experiences RCF or RBF and the RLF parameters shall ideally be set so that the UE and enb will react on the radio problems simultaneously. The standard describes the general procedure how the UE handles radio problems, see figure below. LZT R1A Ericsson AB

132 LTE L13 Radio Network Functionality Scenario Since L12A RRC CONNECTION RE-ESTABLISHMENT REQUEST RRC CONNECTION RE-ESTABLISHMENT RRC CONNECTION RE-ESTABLISHMENT COMPLETE Scenario Before L12A RRC CONNECTION RE-ESTABLISHMENT REQUEST The UE will often experience Radio Link Failure (RLF) when the enb experiences Radio Connection Failure (RCF) or RBF (Radio Bearer Failure) and the RLF parameters shall ideally be set so that the UE and enb will react on the radio problems simultaneously. RRC CONNECTION RE-ESTABLISHMENT REJECT Figure 4-5 Radio Link Failure Signaling Flow in the UE First Phase Second Phase Normal operation Radio problem detection no recovery during T310 no recovery during T311 Return to idle RRC_Connected radio link failure -> RRC Connection Reestablishment Request RRC_IDLE -> New RRC Connection Request Figure 4-6 Radio Link Failure in the UE The UE uses timers T310 and T311 to get time to restore the connection with the enb. During the time T310 + T311 the UE stays in RRC_CONNECTED. If the UE can not reestablish the connection to the enb, the UE switches from RRC_CONNECTED to RRC_IDLE and initiates the procedure to establish a new RRC connection. It may then establish a connection to a new enb Ericsson AB 2012 LZT R1A

133 Radio Connection Supervision The UE uses timers T310 (default 2s) and T311 (default 3s) to get time to restore the connection with the enb. During the time T310 + T311 the UE stays in RRC_CONNECTED. - If the UE cannot reestablish the connection to the enb, the UE switches from RRC_CONNECTED to RRC_IDLE and initiates the procedure to establish a new RRC connection. It may then establish a connection to a new enb. The T310 and T311 timers have preconfigured default values and can only be changed by Ericsson personnel. Figure 4-7 Radio Link Failure Parameters (Ericsson Only) The UE may also, at Radio Link Failure detection, perform an RRC Connection Re-establishment procedure. The enb uses RRC Connection Re-establishment Request as an indication of Radio Link Failure at the UE. The T310 and T311 timers have preconfigured default values and can only be changed by Ericsson personnel. 3.1 Radio Link Monitoring Introduction The UE shall monitor the downlink link quality based on the cell-specific reference signal in order to detect the downlink radio link quality of the serving cell. The UE shall estimate the downlink radio link quality and compare it to the thresholds Q out and Q in for the purpose of monitoring downlink radio link quality of the serving cell. The threshold Q out is defined as the level at which the downlink radio link cannot be reliably received and shall correspond to [10%] block error rate of a hypothetical PDCCH transmission taking into account the PCFICH errors with transmission parameters specified in Table 3-1 Table 3-1 PDCCH/PCFICH transmission parameters for out-of-sync DCI format Attribute Number of control OFDM symbols Value 1A [2]; Bandwidth [10] MHz [3]; [3] MHz Bandwidth [5] MHz LZT R1A Ericsson AB

134 LTE L13 Radio Network Functionality Aggregation level (CCE) Ratio of PDCCH RE energy to average RS RE energy Ratio of PCFICH RE energy to average RS RE energy [4]; Bandwidth = [1.4] MHz 4; Bandwidth = [1.4] MHz 8; Bandwidth [3] MHz [4] db; when single antenna port is used for cellspecific reference signal transmission by the serving cell [1] db: when two or four antenna ports are used for cell-specific reference signal transmission by the serving cell [4] db; when single antenna port is used for cellspecific reference signal transmission by the serving cell [1] db: when two or four antenna ports are used for cell-specific reference signal transmission by the serving cell Note 1: DCI format 1A is defined in section in 3GPP TS [21]. Note 2: A hypothetical PCFICH transmission corresponding to the number of control symbols shall be assumed. The threshold Q in is defined as the level at which the downlink radio link quality can be significantly more reliably received than at Q out and shall correspond to [2%] block error rate of a hypothetical PDCCH transmission taking into account the PCFICH errors with transmission parameters specified in Table 3-2. Table 3-2 PDCCH/PCFICH transmission parameters for in-sync Attribute Value DCI format Number of control OFDM symbols Aggregation level (CCE) 4 Ratio of PDCCH RE energy to average RS RE energy Ratio of PCFICH RE energy to average RS RE energy 1C [2]; Bandwidth [10] MHz [3]; [3] MHz Bandwidth [5] MHz [4]; Bandwidth = [1.4] MHz [0] db; when single antenna port is used for cellspecific reference signal transmission by the serving cell [-3] db; when two or four antenna ports are used for cell-specific reference signal transmission by the serving cell [4] db; when single antenna port is used for cellspecific reference signal transmission by the serving cell [1] db: when two or four antenna ports are used for cell-specific reference signal transmission by the serving cell Note 1: DCI format 1C is defined in section in 3GPP TS [21]. Note 2: A hypothetical PCFICH transmission corresponding to the number of control symbols shall be assumed Ericsson AB 2012 LZT R1A

135 Radio Connection Supervision 3.2 Requirements Minimum requirement when no DRX is used When the downlink radio link quality estimated over the last 200 ms period becomes worse than the threshold Q out, Layer 1 of the UE shall send radio problems indication to its higher layers within 200 ms Q out evaluation period. When the downlink radio link quality estimated over the last 100 ms period becomes better than the threshold Q in, Layer 1 of the UE shall send an in-sync indication to the higher layers within 100 ms Q in evaluation period. The out-of-sync and in-sync evaluations shall be performed. Two successive indications from Layer 1 shall be separated by at least 10 ms. The transmitter power shall be turned off within 40 ms after expiry of T310 timer Minimum requirement when DRX is used When DRX is used the Q out evaluation period (T Evaluate_ Q out_drx ) and the Q in evaluation period (T Evaluate_ Q in_drx ) is specified in Table 3-3 will be used. When the downlink radio link quality estimated over the last T Evaluate_ Q out_drx [s] period becomes worse than the threshold Q out, Layer 1 of the UE shall send outof-sync indication to the higher layers within T Evaluate_ Q out_drx [s] evaluation period. When the downlink radio link quality estimated over the last T Evaluate_ Q in_drx [s] period becomes better than the threshold Q in, Layer 1 of the UE shall send insync indications to the higher layers within T Evaluate_ Q in_drx [s] evaluation period. The out-of-sync and in-sync evaluations shall be performed. Two successive indications from Layer 1 shall be separated by at least max ([10] ms, DRX_cycle_length). Upon start of T310 timer, the UE shall monitor the link for recovery using the evaluation period and Layer 1 indication interval corresponding to the non-drx mode until the expiry of T310 timer. The transmitter power shall be turned off within 40 ms after expiry of T310 counter Minimum requirement at transitions Two successive indications from Layer 1 shall be separated by at least max ([10] ms, DRX_cycle_length). LZT R1A Ericsson AB

136 LTE L13 Radio Network Functionality When the UE transitions between DRX and non-drx or when DRX cycle periodicity changes, for a duration of time equal to the evaluation period corresponding to the second mode after the transition occurs, the UE shall use an evaluation period that is no less than the minimum of evaluation periods corresponding to the first mode and the second mode. Subsequent to this duration, the UE shall use an evaluation period corresponding to the second mode. This requirement shall be applied to both out-of-sync evaluation and insync evaluation. Table 3-3 Q out and Q in Evaluation Period in DRX DRX cycle length (s) T Evaluate_Q out_drx and T Evaluate_Q in_drx (s) (DRX cycles) 0.04 [Note (20)] 0.08 [0.8 (10)] 0.16 [1.6 (10)] 0.32 [3.2 (10)] 0.64 [6.4 (10)] 1.28 [6.4 (5)] 2.56 [12.8 (5)] Note: Evaluation period length in time depends on the length of the DRX cycle in use 4 Principles of radio connection supervision 1. Start: RRC Connected 4. Inactivity timeout 3. RRC Connection Re-establishment Request (received by the enb) 2. RLC Failure Signaling And Payload RB Handling End: RRC Idle Release Figure 4-8 Principles of Radio Connection Supervision 13 RCS starts for a UE and the bearers it uses when the UE enters state RRC_CONNECTED. 14 If an RLC PDU has been retransmitted (ARQ) the maximum number of times the RB experiences RLC Failure. RBF will be indicated Ericsson AB 2012 LZT R1A

137 Radio Connection Supervision 15 If the enb receives RRC Connection Re-establishment Request it will be an indication of RLF at the UE. 16 All DRBs have been inactive too long time. RCS indicates RCF/UE Inactivity to Signaling and Payload Bearer Handling. 4.1 Supervision of UE in RRC_connected state If the UE detects a certain number (N310) of out-of-sync indications, the timer T310 is started. If no recovery (N311 in-sync indications) is detected before T310 expires, RRC Connection Reestablishment is triggered if security is activated, otherwise the UE goes to RRC_IDLE. When the RRC Connection Reestablishment is triggered, the timer T311 is started. If the UE has not found a cell (with a new cell reselection) to send the RRC Connection Reestablishment to, before T311 expires, it releases the resources and goes to RRC_IDLE. If the RBS receives the RRC Connection Reestablishment Request message and if the feature RRC Connection re-establishment is not active, it will send RRC Connection Reestablishment Reject to the UE which will go to RRC_IDLE. If there is a UE context in the cell which the UE triggered RRC Connection Reestablishment to, the RBS will trigger a release of that UE context RLC Failure and PDCCH ordered UL Re-synchronization After a maximum number of ARQ retransmissions (RLC failure), a DL SRB or RB cannot be recovered. After a maximum number of ARQ retransmissions of a DRB or SRB, the RLC shall indicate RBF to the RCS. When enb has detected RLC Failure or PDCCH Ordered UL Re-synchronization failure, the enb will start a wait timer. During this time, the enb will wait for the UE to trigger a Reestablishment to the serving cell. If the Reestablishment is triggered before timer is expired, the enb will execute the Reestablishment procedure. If the timer expires, the enb will trigger a UE release procedure. The T310 and T311 timers have preconfigured default values and can only be changed by Ericsson personnel UE Inactivity RLC monitors UL and DL inactivity per DRB of a UE. If a DRB has been inactive in both uplink and downlink for a certain period, RLC will report inactivity of DRB to RCS. RCS will monitor continuous inactivity of reported DRB for duration equal to at least tinactivitytimer. LZT R1A Ericsson AB

138 LTE L13 Radio Network Functionality If all DRBs of the UE are monitored to be inactive for at least the tinactivitytimer, RCS will trigger a UE release. RCS will abort inactivity supervision if RLC detects activity on a DRB. The three different triggers for UE release are shown below. N310 BLER<Qout RRC Connected T310 BLER<Qout RL Failure -> RRC Connection Reestablishment Request 3 T311 1 RRC Connection Failure Return to idle 4 If ALL DRBs are inactive tinactivitytimer UE Release 2 RLC Failure dlmaxretxthreshold Radio Bearer Failure UE Release Figure 4-9: Supervision of UE in RRC Connected State 5 RRC Connection Re-Establishment This feature was first introduced in L12B. The feature allows the UE to re-establish RRC connection in serving cell in case of radio link failure. From L13A, the feature also allows the UE to re-establish the connection in target cell during handover preparation and execution as well as to an unprepared cell (a cell that does not have the UE context). 5.1 Feature Description The feature minimizes outage time in case lost connection and avoids dropped telephony call in case of successful RRC connection re-establishment. It all starts from the scenario where if a UE experiences a radio link failure, the connection towards RAN and core network needs to be released and setup again. If RRC Connection Reestablishment feature is activated, core network will not notice if the UE has re-established the connection Ericsson AB 2012 LZT R1A

139 Radio Connection Supervision From the on going session, it results in connection interruption time where services such as a VoIP user may need to make a new call that is not desired service wise. It can be expected under this situation an increased load in RAN and core network due to signaling. If a UE experiences a radio link failure, the connection towards RAN and core network needs to be released and setup again Results in connection interruption time A VoIP user may need to make a new call Increased load in RAN and core network due to signaling Figure 4-10: RRC Connection re-establishment -Background/Problem Description A wait timer (not configurable by operator) is started in enodeb when the maximum number of RLC retransmissions has been reached, or maximum number of PDCCH Ordered Re-synchronization failure are detected. This timer is only configurable be Ericsson personnel. During this time, enodeb will wait for UE to trigger RRC Connection Reestablishment. If no RRC Connection Re-establishment Request is received during this time, a UE release will be triggered. The main benefits of this feature is that real time and delay sensitive services such as a VoIP call can survive to a temporary and short period radio link failure. This will help to improve UE drop rate related KPIs in LTE. But in general, it can be said that this feature will shorten the outage in case of radio link failure. This procedure is compared to releasing and setting up a new connection. RRC Connection Re-establishment is: A mechanism for restoring the connection between the UE and an enodeb Triggered by the UE when e.g. radio link failure is experienced A VoIP call can survive a radio link failure Improves UE drop rate KPI Shortens the outage in case of radio link failure Transparent to core network No core network signaling needed when re-establishing a connection (i.e. reduced core network load) Reduced amount of RRC signaling Compared to releasing and setting up a new connection Figure 4-11: Benefits of the Feature LZT R1A Ericsson AB

140 LTE L13 Radio Network Functionality Implementation The new basic feature, RRC Connection Re-establishment introduces a mechanism/procedure for restoring the connection between the UE and an enodeb. This feature requires support for the RRC Connection Re-establishment procedure in the UE. No other network element is impacted by this feature. The procedure is triggered by the UE when, for example, radio link failure is experienced. The UE may also trigger RRC Connection Re-establishment to another (target) cell than the serving cell. RRC Connection Re-establishment is supported: In serving and target cell during handover In an unprepared cell (no UE context) When there is no ongoing UE procedure (E-RAB Setup/Release/Modify, UE Context Setup/Release/Modify etc.), except handover Figure 4-12: Limitations Ericsson AB 2012 LZT R1A

141 Radio Connection Supervision 6 Parameters Parameter tinactivitytimer Description The time a UE can be inactive before it is released. Range: { 0, } Unit: s 86400s = 24 hours Special values: 0 is used to turn off the use of this timer. This means that a UE will not be released due to inactivity. rrcconnreestactive Level: RBS {true, false} LZT R1A Ericsson AB

142

143 Capacity Management Intentionally Blank5 LZT R1A Ericsson AB

144 LTE L13 Radio Network Functionality 5 Power Control, Scheduling and Link Adaptation Objectives After this chapter the participants will be able to: Describe the purpose and use of the function Power Control, Link Adaptation and Scheduling 1. Explain the interaction between Power Control, Link Adaptation and Scheduling 2. Explain Open Loop Power Control for initial access 3. Configure the power for common channels 4. Explain uplink Power Control for PUSCH and PUCCH 5. Explain the impact of TDD 6. Explain impact of MIMO Figure 5-1: Objectives of Chapter 5 Power Control, Link Adaptation and Scheduling Ericsson AB 2012 LZT R1A

145 Capacity Management 1 Introduction Scheduling, Link Adaptation and UL Power Control are three essential functions in the enb. They interact closely and exchange information in order to be able to allocate an appropriate amount of resources, with the right transport format, modulation and coding as well as appropriate UL power at every TTI. The QoS framework also influences the prioritization of logical channels. An interaction overview of these functions is shown in Figure 5-2. Scheduler DRA Link Adaptation SINR HARQ OPP /BLER f MCS MIMO #SBs SINR PUSCH SINR PDSCH Channel Prediction Measure PSD RX N+I Estimate from PHR PSD TX Calculate from CQI: SINR RS GINR SINR PDSCH Calculate G UL DL PSD RX, G UL Power Control CL PC Open Loop RA OL PC Uu BSR/SR (MAC CE / PUCCH or RA) TF UL/DL (PDCCH) RI PMI CFR (PUCCH or PUSCH) CQI PHR (MAC CE) TPC (PDCCH) P 0 (SIB2) Measure PSD RS Estimate SINR RS CQI Figure 5-2 Scheduling, LA and PC Summary LZT R1A Ericsson AB

146 LTE L13 Radio Network Functionality Figure 5-3 lists abbreviations used in Figure 5-2. BSR Buffer Status Report BLER Block Error Ratio CFR Channel Feedback Report CQI Channel Quality Indicator DRA Dynamic Resource Allocation G Gain ( = - Pathloss) HARQ OPP Hybrid Automatic Repeat Request Operating Point LCG Logical Channel Group MCS Modulation Coding Scheme MIMO Multiple Input Multiple Output PHR Power Headroom Report PMI Precoding Matrix Indicator PSD Power Spectrum Density QCI QoS Class Identifier RB Resource Block RI Rank Indicator SB Scehuling Block SINR Signal to Interference and Noise Ratio TF Transport Format TPC Transmit Power Control Figure 5-3 PC, LA & Scheduling Vocabulary A QCI is a scalar that is used as a reference to access node specific parameters that control bearer level packet treatment (e.g. scheduling behavior, queue management thresholds, link layer protocol configurations etc). DRA Dynamic Resource Allocation is another name for Scheduling. CFR Channel Feedback Report consist of CQI, PMI and RI CQI Channel Quality Indicator is an index that is communicated to the enb that indicates link adaptation parameters, Figure 5-4. Code rate is the ratio between the data bits and (the data bits + redundancy). Efficiency is the effective number of bits per symbol Ericsson AB 2012 LZT R1A

147 Capacity Management CQI Index Modulation Code rate x 1024 Efficiency 0 Out of range 1 QPSK QPSK QPSK QPSK QPSK QPSK QAM QAM QAM QAM QAM QAM QAM QAM QAM Figure 5-4 CQI Mapping. PMI Precoding Matrix Indicator is UE s recommendation of the precoding matrix to be used by the system. The number of possible precoding matrixes in the codebook is dependant on the number of antenna ports. PMI is only reported if UE operates in closed loop spatial multiplexing (CLSM) mode. RI Rank Indicator is UE s recommendation for the number of layers i.e. streams to be used in spatial multiplexing. RI is only reported if UE is operating in MIMO mode. PHR Power Headroom Report additional return power available at the UE BSR Buffer Status Report informs the scheduler about the amount of the data to be sent in UL by the UE. MCS Modulation Scheme indicates which modulation type is been used by the scheduler. Possible modulation is QPSK, 16 QAM and 64 QAM. SINR Signal to Interference and Noise Ratio is power ratio between the power of the signal (meaningful information) and power of the interference and the background noise. TPC Transmit Power Control are the control bits used to control the power used in the UL. TF Transport Format is the information sent to the UE that contains the size of the transport block, used modulation and the coding rate. LZT R1A Ericsson AB

148 LTE L13 Radio Network Functionality SB Scheduling Block is the smallest resource unit used by the scheduler, see Figure 5-8 LCG Logical Channel Group is QoS parameter used in UL see Figure 5-6. G power Gain is increase in a power of electromagnetic wave is equal - Pathloss PSD Power Spectrum Density is the power of a signal, divided by the bandwidth. 2 QoS Handling The LTE Quality of Service (QoS) Handling coordinates and assigns the appropriate QoS to other functions in LTE RAN. The RBS maps QCIs (Quality of Service Class Identifiers) to priorities for different Data Radio Bearers (DRBs) in the LTE radio interface and different data flows in the transport network. The LTE QoS Handling complies with the 3GPP QoS concept. It provides service differentiation per UE by support of multiple parallel bearers. To provide service differentiation in the uplink, traffic separation must be ensured between the different data flows within the UE. This is done by offering an operatorconfigurable mapping between QCIs and LCGs (Logical Channel Groups, also sometimes referred to as radio bearer groups). Moreover, service differentiation is enabled by mapping of QCIs to priorities as defined in 3GPP TS In the uplink, these priorities will serve as a basis for the UE to establish differentiation/prioritization between its logical channels. Signaling Radio Bearers (SRBs) are assigned higher priority than Data Radio Bearers (DRBs). SRB1 has higher priority than SRB2. For the UL, the transport network benefits from QoS by mapping QCI to DiffServ Code Point (DSCP) in the RBS. This enables the transport network to prioritize between its different data flows over the S1 interface in the uplink and over the X2 interface for the downlink data in case of Packet Forwarding. For the DL, a similar mapping is performed in the S-GW for the S1 DL data. If a user has multiple bearers with different QCI, these users will be separated in the radio network into different bearers. This separation will achieve benefits for the end-user QoS as it removes the risk that one service such as file download would block the traffic for a voice call. All QoS class identifiers defined by 3GPP are accepted Ericsson AB 2012 LZT R1A

149 parameters QCI, ARP Capacity Management - abspriooverride - priority logicalchannelgroupref - resourcetype dscp DSCP: DiffServ Code Point OSS: Operations Support System Core Network QoS: Quality of Service QCI: QoS Class Indentifier RC: Radio and Core OSS-RC QoS parameters QCI Table, Example of population Standardized QCIs QCI 1 2 : : 9 RT GBR GBR : : Non- GBR Prio 2 4 : : 9 LCG 2 1 : : 3 DSCP : : 12 QCI Table QoS translation QoS Handling paarpoverride QCI table QoS configuration DSCP Priorities LCGs Logical Channel Priorities LCGs Scheduler Grant & Assignm DSCP DL Packet Forwarding (X2) Figure 5-5: QoS Framework UL (S1) Transport Network UE UL/DL (Radio Interface) Radio Network QoS Handling is based on mapping QCIs received from the core network to RBS-specific parameters. This makes it possible to have different priorities and DSCP values. The LTE QoS Handling is realized by a central function in the RBS, which directly influences the radio and transport network behavior. The Scheduler is an essential QoS enabler. In the uplink, the scheduling in the RBS operates on Logical Channel Groups (LCGs) using similar scheduling strategies as in the downlink to grant resources. In uplink, the distribution of the granted resources is done per logical channel internally within the UE using the rate control function. The RBS maps the QCI to LCG and informs the UE about the association of a logical channel to a LCG and the logical channel priority for each logical channel. The reason for using LCGs is that the Buffer Status Report (BSR) is sent per LCG and not per logical channel. This reduces the uplink signaling load. Standardized QCIs (1-9) are used, according to 3GPP TS Non-standardized QCIs (10-256) are all given the same priority, which shall be lower compared to priorities for the standardized QCIs. For the uplink, the priorities are translated to logical channel priorities and sent to the UE, which may differentiate/prioritize between its logical channels. LZT R1A Ericsson AB

150 LTE L13 Radio Network Functionality Mapping QCIs to Logical Channel Groups (LCGs) can be configured in OSS-RC and enables traffic separation in the uplink. There are three LCGs (1-3) available. By default, LCG 1 is assigned to all QCIs. The UE performs prioritization between its own logical channels. This is referred to as UE rate control. An example of this is depicted in Figure 5-6. logicalchannelgroupref QCI LCG Example of mapping of logical channels to logical channel groups in uplink LCID QCI LCG LCG 0 SRB LCG 1 LCG 2 LCG 3 QCI 1 QCI 2 QCI 3 QCI 4 QCI 5 QCI 6 QCI 7 QCI 8 QCI 9 QCI Logical channels BSR 0 BSR 1 BSR 2 BSR BSR 0-3 sent per LCG 0-3 LCG 0 is for SRB UE Figure 5-6. LCG Concept Mapping QCI to DiffServ Code Point (DSCP) for the uplink over S1 and in the downlink for packet forwarding over X2 can be configured in OSS-RC. The DSCP values may also be mapped to Ethernet Priority Bits (P-bits) the values of which are used on the Ethernet layers. The DSCP setting determines the priority for the data stream in the IP transport network. Several QCIs can be mapped to the same DSCP value. Non-standardized QCIs are all given the same configurable DSCP value. Operators with an under provisioned backhaul or backhaul with poor or no QoS handling will be able to control the bandwidth on the transport network by the used of the optional feature Egress traffic management. This is a transport network feature and it is outside the scope of this course. Multiple RBSs can be configured in parallel from OSS-RC. The managed object configuration QciProfilePredefined, which has managed object QciTable as parent, has the configurable parameters priority, dscp, and logicalchannelgroupref. From L12B, LCG 0 can be used also for DRBs Ericsson AB 2012 LZT R1A

151 Capacity Management 3 Scheduling To provide efficient resource usage the LTE concept supports fast scheduling where resources on the shared channels PDSCH and PUSCH are assigned to users and radio bearers on sub-frame basis according to the users momentary traffic demand, QoS requirements and estimated channel quality. This task is done by the uplink (UL) and downlink (DL) schedulers, both situated in the enb. Scheduling is also referred to as Dynamic Resource Allocation (DRA) and is part of the Radio Resource Management (RRM). There are important interactions with other RRM functions such as power control, link adaptation and Inter-cell Interference Control (ICIC). Shared channel transmission Select user and data rate based on instantaneous channel quality Scheduler located in enb Same allocation for all layers in case of MIMO Scheduling in time and frequency domain on SB level (two RBs) Link adaptation in time domain only Time-frequency fading, user #1 data1 data2 Time-frequency data3 data4 fading, user #2 User #1 scheduled User #2 scheduled 1 ms Time Frequency 180 khz Figure 5-7. Channel Dependent Scheduling. In the downlink, the scheduler may assign a set of resource blocks to a user according to a resource allocation scheme while in the uplink resource blocks assigned to a specific user must be contiguous in the frequency to preserve the SC-FDMA structure. Also only a limited set of DFT sizes will be allowed, i.e. multiples of powers of 2, 3 and 5. This further limits the number of RB that can be assigned to a user. In addition to providing an efficient utilization of the radio resource the scheduler is responsible for ensuring the QoS requirements for the individual logical channels to as large extent as possible. When this is not possible due to resource limitations the scheduler performs prioritization between users and logical channels according to the QoS requirements. In the downlink, where the enb has immediate access to the transmit buffers of the radio bearers; the scheduler performs the prioritization both between users and different radio bearers of a user. LZT R1A Ericsson AB

152 LTE L13 Radio Network Functionality In the uplink on the other hand the scheduler only prioritizes between different users based on buffer status reports. The prioritization between different logical channels within one UE will be done in the UE with assistance from the network. Although fast dynamic scheduling is the base line for LTE scheduling, several methods for limiting the control signaling demands for services such as speech - (VoIP) where small packets are generated regularly - have been discussed in 3GPP. A concept where resources are assigned semi-statically called semipersistent scheduling has been agreed. In L13A, a combination of Delay Based Scheduling and Semi Persistent Scheduling can be used for VoIP traffic. Link Adaptation, which includes transport format selection, is closely related to scheduling and the two functions interact by exchanging information prior to each scheduling decision. The operator controls part of the scheduling behavior via the QoS framework. The operator is able to control some of the scheduler behavior in order to meet their needs. This includes the PDCCH and Physical Uplink Control Channel (PUCCH) resources. These adjustments can have a direct impact on user peak throughput and cell capacity. The smallest time/frequency entity that the scheduler may assign consists of twelve sub-carriers (180 khz) in the frequency domain and a sub-frame (1ms) in time. This corresponds to two 180kHz 0.5ms physical resource blocks that are consecutive in time and is referred to as a Scheduling Block (SB). In case of MIMO the resource allocation is the same for all streams. Channel variations can be exploited for multi-user diversity (i.e. scheduling users in constructive fading) both in time and frequency domain See Figure 5-8. One Scheduling Block Two RBs 0.5ms 0.5ms 180 khz f 1 ms Figure 5-8. Scheduling Block. t In the downlink, the resources handled by the scheduler per cell are: Physical Resource Blocks Ericsson AB 2012 LZT R1A

153 Capacity Management PDCCH Resources DL Power (not in this release) TX rank Baseband module processing capability In the uplink, the resources handled by the scheduler per cell are: Physical Resource Blocks Baseband module processing capability PUCCH Resources Common resource is the baseband processing power of the RBS in UL and DL respectively. There are also different licenses for UL and DL UlBasebandCapacity DlBasebandCapacity (Mbps). DL Scheduling: Control Plane, Retransmissions User Plane per radio bearer UL Scheduling UE scheduling, Retransmissions, CQI reports enb Baseband Processing Capability Cell DL Physical Resource Blocks DL Power Cell UL PDCCH Resources Physical Resource Blocks TX Rank PDCCH Resources UL Scheduling Logical Channel Prioritization Figure 5-9: Resources Handled by Scheduler The Scheduler handles the distribution of control and user data on the physical resource blocks in the time and frequency domains across the radio interface. It enables users to be multiplexed and scheduled simultaneously, and facilitates efficient use of spectral and hardware resources for optimization of user throughput and cell capacity. The LTE Scheduler supports resource fair and Minimum Rate proportional Fair (with its 5 variants, described later) scheduling strategies. The Scheduler supports Single User-Multiple Input Multiple Output (SU-MIMO) which improves coverage and peak bit rates. LZT R1A Ericsson AB

154 LTE L13 Radio Network Functionality 3.1 Scheduling Details For every cell in every transmission time interval (1ms), the Scheduler determines the UEs that are assigned resources. Each radio bearer is given a certain scheduling priority, based on algorithms which take the QCI (QoS Class Identifier) related parameters as input. A higher scheduling priority gives the radio bearer a higher probability to obtain resources and enables the UE to perform transmission or reception. The allocation of resources is made per UE. The UE priority is defined as the highest scheduling priority of the radio bearers belonging to the UE, including retransmissions. The UE with the highest priority is selected first for transmission. In the uplink, some radio bearers with lower priorities may get a "free ride" if a radio bearer with a high priority belongs to the same UE. In the uplink, when the UE has many radio bearers, they are grouped into Logical Channel Groups (LCGs, also sometimes referred to as radio bearer groups). The buffer status report (BSR) is then expanded so it reports the buffer status per radio bearer group. Each LCG is given a certain priority (since the buffer status is only visible per LCG in the UL scheduler). The downlink scheduler is similar. The differences are e.g.: no LCG or BSR, no free ride, and that the scheduler queue is common to all UEs on the same baseband module, which can cover more than one cell. The Scheduler controls all radio interface resources, except the following physical signal and channel transmissions, as shown in Figure Reference Signal (RS) Sounding Reference Signal (SRS) Synchronization Channel (SCH) Physical Broadcast Channel (PBCH) Physical HARQ Indicator Channel (PHICH) Physical Uplink Control Channel (PUCCH) Physical Random Access Channel (PRACH) Physical Control Format Indicator Channel (PCFICH) Ericsson AB 2012 LZT R1A

155 Capacity Management The diagram below illustrates the Scheduler Control. Downlink Uplink DL Scheduling UL Scheduling BCCH PCCH DTCH DCCH CCCH DTCH DCCH CCCH MIB SIB Logical Channels type of information (traffic/control) BCH PCH DL-SCH UL-SCH RACH Transport Channels how and with what characteristics (common/shared/mc/bc) -Sched TF DL -Sched grant UL -Pwr Ctrl cmd PDCCH -HARQ info info ACK/NACK -CQI -(N)ACK -SR. PBCH PDSCH PDCCH PCFICH PHICH PUCCH PUSCH RS PSS SSS RS SRS Figure 5-10: Scheduler Control PRACH Physical Channels bits, symbols, modulation, radio frames etc Physical Signals only L1 info The priority of logical channels for both uplink and downlink include the following. Common channels are given the highest priority: System information data on the Broadcast Control Channel (BCCH) Random access messages on the Common Control Channel (CCCH) Paging messages on the Paging Control Channel (PCCH) Link adaptation is done to reach the cell edge with a certain probability of decoding, determined by the link budget. In the downlink, DCCH is given the next priority, with SRB1 having higher priority than SRB2. Finally, the Dedicated Traffic Channel (DTCH) is allocated. Service differentiation does not exist between different DTCH. HARQ retransmissions are given higher priority than new transmissions. In the uplink, HARQ retransmissions are given higher priority than both new data transmissions and DCCH, because synchronized HARQ retransmissions are used. LZT R1A Ericsson AB

156 Priority Order Priority Order LTE L13 Radio Network Functionality The different transmissions are prioritized in the following order: DL Common channels HARQ retransmissions of DCCH Initial transmissions of DCCH HARQ retransmissions of DTCH Initial transmissions of DTCH UL Transmissions of random access msg 3 HARQ retransmissions Initial transmissions of DCCH Initial transmissions of DTCH Figure Scheduling internal priorities. No scheduling grants need to be transmitted for the synchronous retransmissions in the uplink. At the beginning of each transmission time interval, the Scheduler receives information on available resource blocks, and available downlink power in the cell. It may also get an indication of the uplink buffer status from the UE in a Buffer Status Report (BSR). The Scheduler, together with Link Adaptation and power control then assigns an appropriate amount of resources to the UE. This is done provided the UE is in RRC Connected State, the uplink is synchronized and data is in the buffer. The scheduling algorithm for DTCH in both the uplink and downlink depends on the QCI configuration. In L12 two main scheduling strategies can be used: Resource Fair (Round Robin), which selects (given highest priority) the radio bearer with the largest time since last scheduling grant (not to be confused with Delay Scheduling). The UE priority increases with the time elapsed since the last assignment/grant. This is a simple and robust form of scheduling that works well for most services. With Round Robin, the scheduling decisions are mainly delay-based. Resource Fairness is also used among users with same QCI-priority level at strict QCI priority scheduling. The resource fair is the basic scheduling strategy. Minimum Rate Proportional Fair, which can be applied as one of its five variants (Equal Rate, Proportional Fair High, Proportional Fair Medium, Proportional Fair Low, Maximum C/I). It provides a trade-off between user fairness and system performance. By prioritizing users experiencing good channel quality, a higher throughput can be achieved when compared to Resource Fair Ericsson AB 2012 LZT R1A

157 Capacity Management Figure 5-12 below summarizes the available scheduling strategies that can be used: Scheduling Strategy Resource Fair (Round Robin) Description Resource Fair gives more fairness to users served, but leads to lower capacity. Equal Rate Equal rate is the most fair variant of the proportional fair scheduling algorithms. It strives to give users an equal rate. Proportional Fair High Proportional Fair High is a high fairness variant of the proportional fair scheduling algorithm. Proportional Fair Medium Proportional Fair Medium is a medium fairness variant of the proportional fair scheduling algorithm. Proportional Fair Low Maximum C/I Proportional Fair Low is a Low fairness variant of the proportional fair scheduling algorithm. Maximum C/I is the least fair variant of the proportional fair scheduling algorithm. Figure 5-12: Available Scheduling Strategies On the uplink scheduling there is also the possibility that the spectrum allocation for PUSCH transmissions are performed based on frequency-dependant channel knowledge instead of the original Resource Fair strategy. This allows the scheduler not only to prioritize based on time interval, but also to decide which spectrum resources will be used (the part of the spectrum with better quality will be used), leading to improved spectrum efficiency and performance. This functionality is introduced on L12 (Uplink Frequency Selective Scheduling) and will be further explained in a separate section. The operator can configure the number of PUCCH resources for the scheduling request and the Channel Quality Indicator (CQI) to control the trade-off between the number of supported users and the uplink peak throughput. This is done with the two parameters noofpucchcqiusers and noofpucchsrusers. Users can be periodically assigned CQI resources on PUCCH or dynamically on PUSCH. Users that are not assigned SR resources will have to request uplink resources by performing a Random Access. An overview of the inputs and outputs of the Scheduler in uplink and downlink is shown in the figure below. Unless otherwise stated, both uplink and downlink scheduling is considered. The diagram below illustrates scheduler interactions. LZT R1A Ericsson AB

158 parameters Resource Alloc QCI, ARP max TBS LTE L13 Radio Network Functionality OSS-RC QoS parameters QCI table Core NW QCI table QoS translation QoS configuration RRC Connection Request (UL) RB and symbols DL Power PHR (UL) Priorities LCGs Tx rank(dl) MCS(DL) #RBs(UL) Scheduler UE prioritization Resource Allocation PDCCH resources Buffer estimation (UL) Link Adaptation TF selection SINR estimation # CW Output Assignment (DL) Grant (UL) Available RBS resources BB Capacity Throughput (licence) Data related inputs HARQ retransmissions Data buffer (DL) User related inputs UE capabilities UE measurement gaps Sync status SINR(UL) Grants (UL) Assignments (DL) TPC (UL) Power Control UL closed & open loop PC RI(DL), CQI (DL), PMI(DL) SR (UL), BSR (UL) UE Figure 5-13: Scheduler Interactions At the beginning of each Transmission Time Interval (TTI), the scheduler receives a list of PRBs, available power, PDCCH capacity and the number of Orthogonal Frequency Division Multiplexing (OFDM) symbols for user data. Provided there are resources available, the scheduler then checks the baseband capacity. The UE to be scheduled is also validated to ensure it is available for receiving data, and it is synchronized. The packet queues are given priority weights depending on their channel type and QoS requirements. The UE weight then becomes the maximum of its respective packet queue weights. This high level flow for scheduling described above in Figure 5-13 is also applied when Minimum Rate Proportional Fair Scheduler and/or Uplink Frequency Selective Scheduler are active. Note that in the Minimum Rate Proportional Fair Scheduling the schedulers take also into account the estimated channel quality and average rate. In the Uplink Frequency Selective Scheduling case, the scheduler allocates UL resources based on frequency-selective channel quality information. Link Adaptation performs transport format selection to enforce the quality requirement while using the available resources in an efficient way. Transport format selection includes selecting the modulation and coding scheme and transport block size Ericsson AB 2012 LZT R1A

159 Capacity Management In the downlink, the selection is channel dependent because Link Adaptation uses CQI reports from the UE to adapt the transmissions to current radio conditions. For the uplink, Link Adaptation takes SINR into account. The SINR is based on measurements on the uplink demodulation reference signal. The transport format selection influences the scheduling decision, so the scheduling decision is implicitly influenced by channel quality estimations (CQI for downlink and SINR for uplink). Link Adaptation is not used in the uplink for random access message 3 (RRC connection request) or retransmissions. Link Adaptation for downlink common channels uses a default SINR for these transmissions to reach the cell edge. The Scheduler provides link adaptation with RLC and MAC header sizes taken into account when selecting a transport format. The Scheduler retrieves information about the number of scheduling blocks and modulation and coding scheme to use in both uplink and downlink from Link Adaptation. Antenna mapping, part of Link Adaptation, controls multi-antenna transmission by deciding the antenna mapping mode (TX diversity, spatial multiplexing or beamforming, as well as submodes within each mode), spatial multiplexing rank and spatial multiplexing precoding matrix. Channel prediction, also part of Link Adaptation, provides information needed for decisions in the other Link Adaptation functions and Power Control. It includes collecting channel measurements, made in the downlink by the UE and sent to the RBS in channel feedback reports containing CQI, precoding matrix indicator (PMI), and rank indicator (RI). In UL the SINR is considered. LZT R1A Ericsson AB

160 LTE L13 Radio Network Functionality Downlink scheduling framework The overall scheduling concept for the downlink is illustrated in Figure To support fast channel dependent link adaptation and channel dependent time and frequency domain scheduling the UE may be configured to report the Channel Quality Indicator (CQI) reports. Typically, the UE bases the CQI reports on measurements on DL reference signals. For TDD the channel reciprocity could be utilized for channel dependent scheduling to some extent, although one must keep in mind that even if the fading characteristics are reciprocal in TDD the UL and DL will experience different interference. For FDD, there might be an unbalance in path gain between uplink and downlink. Based on the CQI reports and QoS requirements of the different logical channels the scheduler assigns time and frequency resources, i.e. scheduling blocks. The resource assignment is signaled on the Physical Downlink Control Channel (PDCCH). The UE monitors the control channels to determine if it is scheduled on the shared channel (PDSCH) and if so, what physical layer resources to find the data scheduled. Ue provides a Channel Quality Report (CQI) based on DL reference symbol measurements DL Scheduler assigns resources per RB based on QoS, CQI etc. Resource allocation is transmitted in same TTI as data noofpucchcqiusers [0-4000] controls the max no of CQI users on PUCCH per cell DL scheduler & Link Adaptation Data (PDSCH) Resource allocation (PDCCH) Reference symbols CQI report (PUCCH/PUSCH) Figure DL scheduling mechanism. enodeb The DL Scheduler assigns physical resource blocks in the frequency domain, on a subframe/transmission time interval basis. In the downlink, the physical resource blocks are assigned from lower frequencies to higher. Downlink Link Adaptation uses the CQI as channel quality feedback from the UE. The CQI is sent in uplink on PUCCH or PUSCH. The number of CQI users per cell on PUCCH is configurable with the parameter noofpucchcqiusers. UE Ericsson AB 2012 LZT R1A

161 Capacity Management CQI Reporting The CQI reports will be transmitted on the Physical Uplink Control Channel (PUCCH) if there is no Physical Uplink Shared Channel (PUSCH) resource allocated. The PUCCH resources for CQI reporting will be configured through RRC. The resources are autonomously revoked when the UE looses UL synchronization but can also be revoked through RRC. When the UE has simultaneous UL data and has an allocation on the PUSCH, the PUCCH resources for CQI reporting cannot be used due to the single carrier structure. In this case the CQI reports are transmitted on the PUSCH time multiplexed (before DFT precoding) with the data transport block on PUSCH. It is agreed in 3GPP that the enb should not need to blindly detect if there is a CQI report included. This means that the CQI transmissions must be known to the enb and the UE cannot transmit CQI reports autonomously. The size of a PUCCH CQI is very limited, in the order of 10 bits. The PUSCH report on the other hand could allow for more bits and different CQI formats can be used on the PUCCH and on the PUSCH. The enb can configure (through RRC) the UE to transmit CQI report periodically. As illustrated in Figure 5-15, when a periodic CQI reporting instance coincides with an UL transmission on the PUSCH the CQI report will be transmitted on the PUSCH and the PUCCH resources are unused. In addition to the periodic reporting, a-periodic reporting is used. A CQI report can also be requested from the enb either using a poll bit in the grant or to use a specific TF for indicating only CQI. CQI not transmitted on PUCCH as data is transmitted on PUSCH Periodic CQI reporting PUCCH PUSCH CQI transmitted on PUSCH together with UL data Only CQI transmitted Figure CQI transmission. Different CQI formats can be used on the PUCCH and PUSCH. A-periodic CQI reports can be requested by the enb. With a wide-band average CQI is based on an average of the SINR in the mutual information domain over a set of sub-bands. The set of sub-bands, S, is semi-statically configured through higher layer signaling (RRC). The actual averaging is not specified by 3GPP. A wide-band average CQI report could consists of a four bit index to a 16-entry CQI table with approximately equal step size in equivalent SINR. The CQI is defined in terms of channel coding rate and modulation order (QPSK, 16QAM, 64QAM). LZT R1A Ericsson AB

162 LTE L13 Radio Network Functionality In addition to a wide-band average two different schemes for PUSCH transmission have been identified for multi-band reports, a UE-Selected subband CQI and a Higher Layer-configured subband CQI. Ericsson product uses Higher Layer-configured subband CQI. In both cases a wide-band average is computed and used as a reference. In addition, M sub-bands (M could be fixed or configured) are selected and encoded differentially using three bits relative to the wide-band average. In the case UE selected sub-band CQI the UE selects M sub-bands to report. The UE internal procedure to select sub-bands is not specified but the selected sub-bands should correspond to the highest CQI values. The sub-band selection is signaled using, N log 2 M bits where N is the number of sub-bands needed to cover the system bandwidth. The supported sub-band sizes and M values is a function of system bandwidths. The Higher Layer configured sub-bands scheme is similar with the exception that the enb configures what sub-bands the UE should report. Figure 5-16 shows the sub-band sizes and the number of sub-bands as a function of the DL system bandwidth for aperiodic PUSCH reporting. The number of subbands (M) is the M best sub-bands, selected by the UE (UE selected) or selected by the enb (Higher layer configured). In the current release, only higher layer configured sub-band reporting is used. N System BW [RBs] k sub-band size [RBs] (UE selected) M # of subbands (UE selected) k sub-band size [RBs] (Higher layer conf) S # of sub-bands (Higher layer conf) 6-7 N/A N/A N/A N/A N / k N / k N / k N / k Figure CQI aperiodic sub-band reporting on PUSCH Ericsson AB 2012 LZT R1A

163 Capacity Management The CQI measurement methodology is only controlled by the UE. One issue is how to measure the interference and if filtering should be applied. The interference is preferably estimated measured on RE (Resource Element) not corresponding to RS (Reference Symbol) for surrounding cells to capture the actual cell load. If the interference is measured on RS the interference will be overestimated as the RS are always transmitted and do not reflect the load on the data channel. If the interference is instead measured on RE not corresponding to RS the interference will be dependent on if there is data transmission or not and reflect the actual load. In this case the interference might be bursty at low to medium loads. Even with full load the interference will vary rapidly with instantaneous pre-coding. The rapidly varying interference has proven to be a serious problem in simulations both for fast link adaptation and scheduling and results in very high HARQ BLER with loss of peak rates and efficiency. Interference filtering has proven to be an efficient tool for handling rapid interference variations. No enb configuration of the CQI measurement processing is available in LTE Uplink scheduling framework Uplink scheduling takes place in the RBS, and resources are assigned per UE. When UE has several radio bearers, the UE performs the prioritization between the radio bearers and is referred to as the UE rate control function. The basic UL scheduling concept is illustrated in Figure 5-17 and Figure The UE informs the scheduler when data arrives in the transmit buffer with a Scheduling Request (SR). The scheduler selects the time/frequency resources the UE shall use. With support from the link adaptation function also the transport block size, modulation, coding and antenna scheme is selected, i.e. the link adaptation is performed in the enb. The selected transport format is signaled together with information on the user ID to the UE. This means that the UE is mandated to use a certain transport format and that the enb is already aware of the transmission parameters when detecting the UL data transmission. As a consequence there is no need for an UL control channel to inform the enb (E-TFCI in WCDMA). This reduces the amount of control signaling required in the uplink, which is important from a coverage perspective, especially with the short 1 ms TTI.. The diagram below illustrates the UL scheduling mechanism. LZT R1A Ericsson AB

164 CQI Status TF selection LTE L13 Radio Network Functionality UE requests UL transmission via scheduling request Scheduler assigns initial resources without detailed knowledge of buffer content More detailed buffer status report may follow in connection with data Either D-SR on PUCCH or RA-SR on RACH Measurements (SINR) UL Scheduler Link Adaptation & Power Ctrl noofpucchsrusers enodeb Demodulation RS Scheduling Request(PUCCH/RACH) Resource assignment (PDCCH) BSR (PUSCH) PHR (PUSCH) Data (PUSCH) UE Figure UL Scheduling Mechanism. The scheduling request uses PUCCH format 1 and the number of scheduling request users per cell is configurable with an operator controlled parameter noofpucchsrusers which determines the number of scheduling request resources available on PUCCH. The periodicity of the scheduling request is 10 ms, and equal for all UE in the cell. If no scheduling request resources are allocated on PUCCH for the UE, the UE uses the random access process to request resources. Uplink transport format controlled by NodeB No TFC selection in the UE enodeb enodeb Buffer Buffer Scheduler Multiplexing Scheduler Uplink channel quality Modulation, coding UE UE Modulation, coding Downlink channel quality Priority handling Multiplexing Buffer Buffer Downlink Uplink Figure UL Scheduling Ericsson AB 2012 LZT R1A

165 Capacity Management The assigned resources and transmission parameters is revealed to the UE via the PDCCH. Additional Scheduling Information (SI) such as Buffer Status Report (BSR) or Power Headroom Report (PHR) may be transmitted together with data as MAC control elements. The enb may configure the UE to transmit a wideband Sounding Reference Signal (SRS) that can be used for estimating the UL channel quality. Additional channel quality estimates can be obtained from other UL transmissions such as, data transmission or control signaling (CQI reports and HARQ ACK/NACK signals). Scheduling resources among users in the uplink is complicated by the fact that the scheduler is situated in the enb and is not automatically aware of the users resource demand, i.e., what users and what radio bearers have data to transmit and how much data there is in the transmit buffers. The concept for uplink scheduling suggested is based on a resource reservation principle. When data arrives to the transmit buffers of a UE and the UE has no grant for transmission on the PUSCH, the UE needs to request permission to transmit by sending a Scheduling Request (SR). The SR will either be transmitted on the RACH channel (RA-SR) or on dedicated resources on the PUCCH (Dedicated SR, D-SR) if such resources are available. The PUCCH resources for dedicated SR are assigned and revoked by the enb through RRC. In addition, the resources are autonomously revoked when the UE looses UL synchronization. The UL scheduler monitors the users requests and distributes the available resources among the various users. A Dedicated Scheduling Request (D-SR) is typically used when the UE uplink is time synchronized. The purpose is to enable UE to rapidly request resources for uplink data transmission. Each active user is assigned a dedicated channel for performing the scheduling request. The benefit with this method is that no UE ID has to be transmitted explicitly, since the UE is identified by the channel used. Furthermore, no intra-cell collisions will occur in contrast to the contention based approach. The D-SR is repeatedly transmitted on consecutive SR opportunities on PUCCH until the UE receives an UL grant on PDCCH. The transmission is stopped when; PUCCH resources are released, a maximum number of D-SR has been transmitted and/or UL synch is lost, even if the UE has not received any UL grant on PDCCH. After stopping transmission on the D-SR, the UE transmits on the RA-SR (i.e. accesses the system via RACH). The Random Access Scheduling Request (RA-SR) is used when the UE has lost UL synchronization or if it has no D-SR resources. From a scheduling request the scheduler has limited knowledge of what type of data and of what priority the UE has. For further information a grant is issued by the scheduler. The grant addresses a UE and not a specific logical channel. In its simplest form the scheduling grant is valid only for the next UL TTI. LZT R1A Ericsson AB

166 LTE L13 Radio Network Functionality Together with the uplink data, the UE transmits buffer status information to the RBS such that the uplink scheduler knows roughly the amount of data in the UE buffer. In cooperation with the link adaptation and power control functionality the uplink scheduler uses this information together with the channel quality information to assign an appropriate number of resource blocks to UE in the uplink. The more detained Buffer Status Report (BSR) is provided by the UE to the enb as a MAC control element. Additional information such as UE power headroom is also used to support link adaptation and power control. A BSR is triggered when at least one of the following criteria is fulfilled: UL data arrives in the UE transmission buffer and the data belongs to a radio bearer (logical channel) group with higher priority than those for which data already existed in the UE transmission buffer. when UL-SCH resources are allocated and number of padding bits is larger than the BSR size when the UE arrives to a new cell the periodic BSR timer expires A triggered BSR is cancelled in case the uplink grant is large enough to accommodate all pending data, but not both data and BSR. A SR is triggered and transmitted in addition if the BSR is not transmitted before the first available SR opportunity. The resources not used by the physical signals PUCCH and PRACH are available for uplink scheduling, so PUCCH and PRACH have absolute priority over other control and user data channels in the uplink. The random access response message on PDSCH contains the UL grant for random access message 3 (RRC connection request). Uplink grants are produced by the UL Scheduler and sent to the UE on PDCCH. The UE has to follow this grant, which means that there is no TFC (Transport Format Combination) selection in the UE, as for E-UL in WCDMA. When the 'ICIC - autonomous resource allocation' feature is activated in uplink with parameter ulinterferencemanagementactive, a random selection is made as to the allocation shall start at the lower or the upper part of the spectrum. It is possible to configure a proportion of the UL bandwidth to be used by the UL scheduler using the parameter ulfrequencyallocationproportion. This parameter only takes effect when ulinterferencemanagementactive is set to true. Be aware that using only a fraction of the UL bandwidth could affect peak throughput and capacity in a negative way Ericsson AB 2012 LZT R1A

167 Capacity Management The diagram below illustrates UL scheduling Allocation (ICIC). Random starting point (upper/lower) f t ulinterferencemanagementactive [true/false] ulfrequencyallocationproportion = [ ] Figure 5-19: UL Scheduling Allocation (ICIC) To increase capacity and coverage, Time Spread Allocation Scheduling is used in uplink. This enables multiple UE to be scheduled in one subframe by distributing one UE transmission over several subframes, but at a lower bandwidth. This leads to an improved link budget and improved capacity and coverage. See Figure Without Time Spread Allocation User 1 User 2 User 3 f t With Time Spread Allocation f User 1 User 2 User 3 Figure 5-20: UL Scheduling Allocation t LZT R1A Ericsson AB

168 LTE L13 Radio Network Functionality UE rate control The Logical Channel Prioritization procedure is applied when a new transmission is performed. RRC controls the scheduling of uplink data by signalling for each logical channel: priority where an increasing priority value indicates a lower priority level, prioritisedbitrate (e.g bytes per sec) which sets the Prioritized Bit Rate (PBR), bucketsizeduration (ms) which sets the Bucket Size Duration (BSD). In L12 prioritisedbitrate is set to infinity and that means that all the buckets are filled up and the only prioritization is the logical channel id. The general purpose of PBR is to avoid starvation. Decreasing priority order Bj B3 B2 B1 TTI n+3 TTI n+2 TTI n+1 TTI n... Max Bucket size= PBR x BSD At each TTI Bucket size = bucket size + PBR DTCH... DTCH DCCH DCCH PBR LCH4 PBR LCH3 PBR LCH2 UL-SCH PBR LCH1 Figure 5-21: Logical Channel Prioritzation UL The UE maintains a variable Bj for each logical channel j. Bj is initialized to zero when the related logical channel is established. It is incremented by the product PBR TTI duration for each TTI, where PBR is Prioritized Bit Rate of logical channel j. However, the value of Bj can never exceed the bucket size and if the value of Bj is larger than the bucket size of logical channel j, it is set to the bucket size. The bucket size of a logical channel is equal to PBR BSD, where PBR and BSD are configured by upper layers Ericsson AB 2012 LZT R1A

169 Capacity Management The UE performs the following Logical Channel Prioritization procedure when a new transmission is performed: -The UE allocates resources to the logical channels in the following steps: 17 All the logical channels with Bj > 0 are allocated resources in a decreasing priority order. If the PBR of a radio bearer is set to infinity, the UE allocates resources for all the data that is available for transmission on the radio bearer before meeting the PBR of the lower priority radio bearer(s); 18 the UE decrements Bj by the total size of MAC SDUs served to logical channel j in Step 1 NOTE: The value of Bj can be negative. 19 if any resources remain, all the logical channels are served in a strict decreasing priority order (regardless of the value of Bj) until either the data for that logical channel or the UL grant is exhausted, whichever comes first. Logical channels configured with equal priority are served equally. 3.2 QoS Aware Scheduler The QoS-Aware Scheduler feature allows configuration of the scheduler on a QoS Class Indictor (QCI) basis. This can be used to configure absolute priority scheduling for Data Radio Bearers (DRBs) in regard to air interface resources. Absolute priority for DRBs must be applied conservatively. It is recommended for use for prioritization of QCI1 (Voice over Internet Protocol (VoIP)) and QCI5 (IMS signaling) only. It is not recommended to be used for non-guaranteed Bit Rate (GBR) traffic unless it corresponds to infrequent and low-intensity traffic similar to signaling. A table showing the standardized QCIs according to 3GPP TS is shown below: LZT R1A Ericsson AB

170 LTE L13 Radio Network Functionality QCI Resource Type Priority Packet Delay Budget (NOTE 1) Packet Error Loss Rate (NOTE 2) Example Services ms 10-2 Conversational Voice (NOTE 3) ms 10-3 Conversational Video (Live Streaming) (NOTE 3) GBR ms 10-3 Real Time Gaming (NOTE 3) ms 10-6 Non-Conversational Video (Buffered (NOTE 3) Streaming) ms 10-6 IMS Signalling (NOTE 3) 6 Video (Buffered Streaming) (NOTE 4) ms 10-6 TCP-based (e.g., www, , chat, ftp, p2p file sharing, progressive video, etc.) 7 Non-GBR Voice, (NOTE 3) ms 10-3 Video (Live Streaming) Interactive Gaming 8 (NOTE 5) ms 10-6 TCP-based (e.g., www, , chat, ftp, p2p file Video (Buffered Streaming) 9 9 sharing, progressive video, etc.) (NOTE 6) NOTE 1: A delay of 20 ms for the delay between a PCEF and a radio base station should be subtracted from a given PDB to derive the packet delay budget that applies to the radio interface. This delay is the average between the case where the PCEF is located "close" to the radio base station (roughly 10 ms) and the case where the PCEF is located "far" from the radio base station, e.g. in case of roaming with home routed traffic (the one-way packet delay between Europe and the US west coast is roughly 50 ms). The average takes into account that roaming is a less typical scenario. It is expected that subtracting this average delay of 20 ms from a given PDB will lead to desired end-to-end performance in most typical cases. Also, note that the PDB defines an upper bound. Actual packet delays - in particular for GBR traffic - should typically be lower than the PDB specified for a QCI as long as the UE has sufficient radio channel quality. NOTE 2: The rate of non congestion related packet losses that may occur between a radio base station and a PCEF should be regarded to be negligible. A PELR value specified for a standardized QCI therefore applies completely to the radio interface between a UE and radio base station. NOTE 3: This QCI is typically associated with an operator controlled service, i.e., a service where the SDF aggregate's uplink / downlink packet filters are known at the point in time when the SDF aggregate is authorized. In case of E-UTRAN this is the point in time when a corresponding dedicated EPS bearer is established / modified. NOTE 4: This QCI could be used for prioritization of specific services according to operator configuration. NOTE 5: This QCI could be used for a dedicated "premium bearer" (e.g. associated with premium content) for any subscriber / subscriber group. Also in this case, the SDF aggregate's uplink / downlink packet filters are known at the point in time when the SDF aggregate is authorized. Alternatively, this QCI could be used for the default bearer of a UE/PDN for "premium subscribers". NOTE 6: This QCI is typically used for the default bearer of a UE/PDN for non privileged subscribers. Note that AMBR can be used as a "tool" to provide subscriber differentiation between subscriber groups connected to the same PDN with the same QCI on the default bearer Ericsson AB 2012 LZT R1A

171 Capacity Management QCIs are assigned priorities using the QciProfilePredefined.priority parameter. The default values for the QciProfilePredefined.priority parameter are according to 3GPP Next, QCIs are grouped into Logical Channel Groups (LCGs) by assigning the QciProfilePredefined.logicalChannelGroupRef parameter to point to a certain instance of LogicalChannelGroup. abspriooverride= HI_PRIO_OVERRIDE(1) UE Rate Control LCG 1 (eg. QCI 1, 5) abspriooverride= NO_OVERRIDE(0) LCG2 (eg. QCI 2,3,4,6) abspriooverride= NO_OVERRIDE(0) LCG3 (eg. QCI 7,8,9) abspriooverride= NO_OVERRIDE(0) LCG3 (eg. QCI 7,8,9) Scheduled before all other QCIs * ** Round Robin *highest prio represents the LCG **highest prio represents the UE Scheduled according to scheduling algorithm Ericsson AB 2009 Ericsson Internal X (X) Date Figure 5-22 QoS Aware Scheduler Absolute priority is enabled on an LCG basis by means of LogicalChannelGroup.absPrioOverride. Although the configuration is expressed on an LCG basis it applies to both downlink and uplink. QCIs belonging to a LCG with the LogicalChannelGroup.absPrioOverride parameter set to NO_OVERRIDE are mapped to a default priority level lower (less important) than any other DRB priority. This applies to both downlink and uplink scheduling. For QCIs belonging to a LCG with LogicalChannelGroup.absPrioOverride parameter set to HI_PRIO_OVERRIDE, Downlink (DL) scheduling considers the QciProfilePredefined.priority parameter of the DRB as given by the QCI. In uplink, scheduling considers LCGs. In a LCG with LogicalChannelGroup.absPrioOverride assigned HI_PRIO_OVERRIDE, the DRB (or the associated logical channel) with the most important priority is selected to represent the group. The diagram below illustrates The diagram below illustrates the Absolute Priority. LZT R1A Ericsson AB

172 LTE L13 Radio Network Functionality prio: 1 prio: 2 prio:.. Absolute Priority to LCGs set to abspriooverride = HIGH_PRIO_OVERRIDE prio:.. prio: 14 prio: 15 delay: 9 delay: 8 Figure 5-23: Absolute Priority Relative resource sharing for LCGs set to abspriooverride = NO_OVERRIDE using SchedulingStrategy : delay.. delay:.. delay: 1 delay: 0 + CQI priority maximum delay average rate delay air rate retransmission Scheduling Weight 3.3 Minimum Rate Proportional Fair Scheduler As mentioned earlier, the two main scheduling strategies are Resource Fair and Minimum Rate Proportional Fair. The later one is introduced in L12 as an optional feature that works together with the QoS aware scheduler feature. This scheduling strategy is based on channel quality information and scheduled data rates. It allows better sharing of RAN, RBS and baseband resources between different radio bearers, thus providing a trade-off between user demands and system performance. Proportional Fair Scheduling Scheduling decisions are based on a weighted measure of channel quality and average rate Configurable fairness, i.e. influence of channel quality versus average rate Minimum Bit Rate A configured minimum bit rate is maintained by the scheduler UE specific channel quality Previous assignments Buffer sizes Scheduler DL assignments and UL transmission grants Figure 5-24: Minimum Bit Rate Proportional Fair Scheduling - Overview Ericsson AB 2012 LZT R1A

173 Capacity Management Considering the fact that on a mobile network environment, different users will experience different radio conditions at a certain time; this scheduling strategy prioritizes users experiencing good radio quality, thus leading to a higher throughput when compared to the basic scheduling method (Resource Fair). However, to avoid some users being allocated too few or no resources due to poor channel quality, fairness is provided also by taking into account the average rate in scheduler prioritization. The scheduler needs as an input the channel quality of the users in the cell and the average rates of their flows. Therefore it monitors the Channel state Information (CSI) from the UE and measurements performed in the RBS. The average rate is calculated based on previous transmissions for the flows. When resources are divided among users the prioritization is based on a weighted measure of the channel quality and average rate. Proportional fair scheduling takes advantage of that UEs experience good channel quality at different times Schedule UE 1 in t = 1 Schedule UE 2 in t = 2 Schedule UE 3 in t = 3 CQ t UE 3 UE 2 UE 1 Avg. Rate t t=1 t=2 t=3 Figure 5-25: Minimum Bit Rate Proportional Fair Scheduling - Example In many scenarios, cell capacity can be increased by using the low fairness versions of the proportional fair scheduling algorithms. These algorithms increase the share of the resources given to users with good channel conditions. This leads to an overall increase in capacity. The feature also helps in giving a consistent cell edge throughput with the minimum rate Quality of Service (QoS) characteristic. The basic concept is that flows experiencing a bit rate lower than their configured minimum bit rate will be prioritized before flows that have a bit rate above their configured minimum rate. As mentioned above, Proportional Fair scheduling feature takes into accounts both scheduled data rate and radio channel quality. The trade off between user fairness and the system performance can be tuned by Channel Quality Fraction (CQF). The CQF controls how big portion of channel quality should contribute to a user's priority. With increasing fairness the scheduler will try to bring all the users to the same received data rate range. On the other hand system performance could degrade, as more resources need to be allocated to users in bad channel conditions to give them as high data rate as users in good channel conditions. In the most unfair setting (Max C/I), then resources are spent on good channel users yielding a high system throughput. However, cell edge users are disfavored. LZT R1A Ericsson AB

174 LTE L13 Radio Network Functionality To control the CQF s impact on the scheduler, five different scheduling algorithms can be configured, see Figure Proportional Fair Scheduling algorithms parameter: QciProfilePredefined.schedulingAlgorithm Equal Rate Proportional Fair High Proportional Fair Medium Proportional Fair Low Max C/I Min rate in UL/DL. Parameters: QciProfilePredefined.ulMinBitRate QciProfilePredefined.dlMinBitRate Figure 5-26: Configuration These 5 different types of configuration profiles will allow the operator to set the network behavior to a scenario that varies from providing higher fairness to the users (equal rate) to a higher capacity (Max C/I), which provides a higher throughput. This is detailed on the charts of the below Figure 5-27: Opportunistic scheduling enables increased capacity and increased cell edge throughput in higher load scenarios User Throughput Average Cell Throughput Higher fairness Higher capacity Higher fairness Higher capacity Figure 5-27: Benefits Ericsson AB 2012 LZT R1A

175 Capacity Management Minimum rate scheduling To improve fairness between users and also help in giving a consistent cell edge throughput, a minimum rate QoS characteristic is introduced. The basic concept is that flows that experience a bit rate lower than their configured minimum bit rate will have absolute priority over flows in the same QoS class that have a bit rate above their configured minimum rate. In case all users within a QoS class have a bit rate above the minimum bit rate, there will be no difference between the basic proportional scheduling and proportional fair scheduling with minimum rate. The same also applies when all flows are below the minimum rate assuming that they all belong to the same QoS class. Note the difference between minimum bit rate and guaranteed bit rate (GBR): With GBR, the whole system has to make sure that the GBR can be achieved, for example via admission control while the minimum bit rate is a "soft" requirement which only the scheduler strives to fulfill. Note that setting a high value of the minimum bit rate or/and enabling minimum bit rate for too many users, may have a large negative impact on bearers configured without a minimum bit rate. By having the minimum bit rate as a QCI characteristic attribute it also gives a possibility to use it for service or subscription differentiation. For example, a bearer used by premium subscribers might use a higher minimum bit rate and a higher priority. The high priority ensures that the satisfying the bearer's minimum bit rate has higher priority than satisfying the minimum bit rate of bearers with lower priority. 3.4 Uplink Frequency Selective scheduling On previous releases, the UL resource allocation was made only based on timeinterval criteria (Resource Fair). L12 release has brought the possibility to allocate PRBs to UEs also based on frequency-selective channel quality information carried over the SRS (Sounding Reference Signals). This feature is called Uplink Frequency Selective scheduling. The scheduler uses measurements of the channel conditions as well as internal measurements in order to determine which part of the radio resource should be used for each connection. Different traffic characteristics calls for different scheduling strategies: For broadband services where there is large amount of data to each user a scheduling strategy that can utilize the full available bandwidth is beneficial. For narrow band services, such as voice services or small data volume data connections other strategies can be used. LZT R1A Ericsson AB

176 LTE L13 Radio Network Functionality Frequency selective channel quality estimates influence the allocation of transmission grants to UEs UEs transmit in parts of the frequency band where their channel quality is best SRS gain measurements PUSCH gain measurements Channel Quality Estimation CQ Link Adaptation Interference measurements CQ Transmission grants Previous assignments Scheduler Buffer estimation Figure 5-28: Uplink Frequency Selective Scheduling (1/2) UEs that are valid for scheduling Highest weight UEs that have PDCCH resource Validate UEs for scheduling UE 1 UE 2 UE 3 UE 4 UE 5 UE 6 UE 7 UE K Time Domain Scheduling UE 1 UE 2 UE 3 UE 4 Resource Allocation UL synch Data to transmit No DRX... Select users to be scheduled according to the configured scheduling strategy RESOURCE_FAIR PROPORTIONAL_FAIR DELAY_BASED... Select where to schedule UEs according to resource allocation strategy RESOURCE_FAIR FREQUENCY_SELECTIVE UE 2 UE 3 UE 4 UE 1 Figure 5-29: Uplink Frequency Selective Scheduling (2/2) By allocating PUSCH resources to the UEs over the best frequencies, higher spectrum efficiency will be reached, thus improving the cell-edge bit rate and cell capacity (UEs will be given frequencies to transmit with better radio conditions on a certain place and time) Ericsson AB 2012 LZT R1A

177 Capacity Management The UL scheduler tries to frequency multiplex several UEs in the same TTI, even though the UEs might be able to utilize the whole bandwidth. This new functionality is also suitable when there are many simultaneous users which are not able to fully utilize the available bandwidth. In this case each user will be allocated a subset of the available frequency resources. The part of the frequency band that is most favorable depends on the radio propagation and it is unique to each of the users. The channel conditions are constantly changing which means that the measurements need to be updated regularly. Improves cell edge bitrates and cell capacity by using the frequency selectivity of the radio channel Larger transport block size can be used when UEs are scheduled on higher SNR parts of the bandwidth Multi-user diversity gain, i.e. some UEs experience good channel quality when and where others are not UE1 Channel Profile UE2 Channel Profile UE1 Allocation UE2 Allocation Figure 5-30: Benefits Uplink Frequency Selective Scheduling Uplink frequency-selective scheduling means that the spectrum allocation for PUSCH (Physical Uplink Shared Channel) transmissions will be performed based on frequency-dependent channel knowledge. The frequency awareness introduced in Uplink Frequency-Selective Scheduling will allow the scheduler not only to prioritize which users should be scheduled during a certain time interval, but to also decide which spectrum resources should be allocated to each one of these users. As such, resource allocation decisions are made by taking both time and frequency dimensions into account. By performing uplink frequency-selective scheduling, the uplink scheduler attempts to maximize the spectral efficiency. This is done through allocating each UE into a part of the spectrum where the channel quality is good, according to certain measurements. LZT R1A Ericsson AB

178 LTE L13 Radio Network Functionality Sounding The radio channel measurements are based on Sounding Reference Signals (SRS). The sounding reference signals allow the scheduler in the enb to estimate the channel quality of the uplink channel for different UEs in order to be able to apply uplink channel dependent scheduling. The sounding reference signals can also be used to estimate the timing of UE transmissions and to derive timingcontrol commands for uplink time alignment. Periodically transmitted wideband uplink reference signals Provides channel estimates independent of scheduled data transmissions Gain PRB Figure 5-31: What is sounding? Sounding reference signals are transmitted independently of the transmission of any uplink data, i.e. a UE may transmit a sounding reference signal also in subframes where the UE does not have any data transmission. So, the Sounding Reference Signal is not limited to the bandwidth that each UE is scheduled on PUSCH, but can be transmitted on order from the enb on almost any part of the system bandwidth. Therefore the SRS is able to provide channel quality information beyond the bandwidth scheduled for PUSCH. This way SRS is an important element for the frequency selective scheduling. During SRS transmissions, frequency-selective uplink measurements of the signal power will be made. These power measurements will continuously be reported to the uplink scheduler, which uses these measurements in order to estimate the frequency-selective uplink gain of each UE. The estimated gain, together with the distribution of noise and interference within the serving cell, is used by the uplink scheduler when distributing PRB resources to the UEs targeted for scheduling Ericsson AB 2012 LZT R1A

179 Capacity Management SRS transmissions are performed in the SC-FDMA (Single Carrier Frequency Division Multiple Access) symbol, next to DMRS, in subframes configured by the enb. It can be transmitted alone in one subframe or together with PUSCH, as exemplified on Figure 5-32 below: Transmitted standalone (when no PUSCH resource assigned) One subframe... or together with PUSCH ( stealing one PUSCH symbol) SRS bandwidth independent of PUSCH bandwidth Sounding in last symbol Figure 5-32: Sounding Reference Signal Sounding-reference-signal transmissions are configured with a certain bandwidth, indicating how many resource blocks (in the frequency domain) the sounding reference signal will cover a certain period, indicating the distance, in time, between consecutive SRS transmissions (1, 2, 5, 10 subframes) a certain duration, indicating how many times the sounding reference signal will be transmitted The SRS can be allocated per cell and per QCI up to every 5ms. The Uplink Frequency-Selective Scheduling feature enables the operator to select if an SRS should be allocated or not per QCI. The enodeb will only attempt to allocate an SRS resource to UEs that have at least one Logical Channel Group (LCG) that trigger a request for an SRS resource. An LCG triggers request for an SRS resource when the Quality of Service Class Indicator (QCI) of its highest priority configured logical channel has been configured with the parameter srsallocationstrategy set to ACTIVATED. Thus, the preference of a UE to request an SRS resource may change both when adding and when removing a radio bearer. LZT R1A Ericsson AB

180 LTE L13 Radio Network Functionality SRS configuration options can be seen below. Cell specific configuration SRS allocated in every 5:th subframe TDD SRS is allocated only on UpPTS enable/disable: ulsrsenable in EUtranCellFDD UE specific configuration enable/disable per QCI: srsallocationstrategy in QciProfilePredefined Sounding bandwidth and periodicity selected based on system bandwidth System bandwidth [MHz] System bandwidth [PRBs] SRS bandwidth [PRBs] SRS period [ms] Figure 5-33: SRS Configurations Configuration and parameters The scheduler applies frequency-selective resource allocation for PUSCH data associated with an LCG whenever the QCI of the highest priority logical channel in the LCG is set to use resourceallocationstrategy FREQUENCY_SELECTIVE. When applying frequency-selective resource allocation for a UE that has been allocated an SRS resource, the scheduler uses measurements from SRS together with uplink DMRS and interference measurements to allocate the UE to the frequencies with the most favorable signal quality. When applying frequency-selective resource allocation for UEs that have not been allocated an SRS resource, the scheduler bases the scheduling decision on frequency averaged measurements from the uplink DMRS, however taking frequency-selectivity in interference measurements into account. Detailed configuration procedures and parameter information are out of the scope of this course. Nevertheless further information can be found on CPI. Figure 5-34 describes in a simplified way how to enable this feature followed by a table with the related parameters Ericsson AB 2012 LZT R1A

181 Capacity Management The diagram below illustrates Configuraation of UL Frequency Selection Scheduling. Configuration Parameter resourceallocationstrategy in class QciProfilePredefined "Frequency Selective" Scheduler allocates the UEs best resources in terms of channel quality "Resource Fair" Resources are allocated without UE channel quality preference Note that configuration of resource allocation strategy is not dependent on scheduling algorithm Figure 5-34: Configuration of UL Frequency Selective Scheduling 3.5 Uplink end-user bitrate shaping In order to enforce the uplink AMBR (Aggregated Maximum Bit-Rate), the End user bit rate shaping feature is used. This feature acts as a Quality of Service (QoS) enabler and allows for tiered subscription model. The End-User Bitrate Shaping feature enables the operators to provide service differentiation. The MBR is often the key differentiator together with price when defining marketing offerings. The feature also limits the throughput for those users exceeding their subscripted data volume limit, whenever there is the appropriate CN support. The End-User Bitrate Shaping feature takes the Uplink (UL) Aggregated Maximum Bitrate (AMBR) for the non-guaranteed Bitrate (GBR) bearer and the UL Maximum Bitrate (MBR) for each GBR bearer as input. Therefore, in the UL it is possible to shape the traffic to the sum of the maximum aggregated bit-rate for a specific User Equipment (UE) (AMBR) and the MBR for each GBR bearer of that UE. The UE-AMBR and the MBRs associated with each bearer are received over the S1 Application Protocol (S1AP) interface from the Mobility Management Entity (MME) and X2 Application Protocol (X2AP) interface from the enodeb. The granularity of the rate controlled by the MME is 8 kbps. The basic concept of bit rate shaping is the process of delaying packets in a traffic stream to cause it to conform to some previously defined traffic profile. The reasons for this could be: To smoothen traffic entering a network; To enable tiered subscriptions; To discourage cheating, e.g., users upgrade the VoIP codec rate beyond what has been authorized by the network. LZT R1A Ericsson AB

182 LTE L13 Radio Network Functionality This functionality is located on the scheduler (thus on enb) and actuates on the bit-rate of the data carried between the MAC and RLC layers based on inputs provided by the MME via the S1 Application Protocol (S1AP). Those inputs will consist on the UL UE-AMBR and MBR, as depicted by the Figure 5-35 below. Packet Source Packet Sink PDCP PDCP RLC RLC Allow to grant (maxrstbs) UL Validator Rate Shaper UL Scheduler UL UE-AMBR, UL MBR MME MAC MAC enb BSR UE Packet Sink Packet Source Figure 5-35: The UL End-user bit rate shaping in enb Note that the rate shaper is valid only for the uplink direction in L12, so the data is buffered at the UE. For the UL the shaping is impacting the scheduling grants that are given to the UE by the scheduler. The grants are given per UE so the rate that is enforced (r) is effectively the UE-AMBR (applicable to all NGBR bearers) plus the sum of the MBR for all the GBR bearers established for a UE. The traffic exceeding the enforced maximum bit rate will be delayed (buffered) before transmission over the air interface using a Token Bucket algorithm. The basis for this algorithm is that tokens, representing the data volume, are injected at a constant token rate. The Token Bucket algorithm will define the average bit rate r. The next picture summarizes the main characteristics of this algorithm. For a clearer understanding, try to visualize Figure 5-36 inside Figure 5-35, where it is labeled Rate Shaper UL The diagram below illustrates the token bucket algorithm in enb Ericsson AB 2012 LZT R1A

183 Capacity Management Uplink only in L12A; The token bucket algorithm acts on a group of bearers; Use of borrowed tokens so that unnecessary segmentations can be avoided; The maximum delay allowed is up to 25 ms; The bucket size is set to r times; Algorithm parameters Token rate, r Bucket size, b Non- GBR GBR SRB Token rate, r Token bucket Bucket size, b R Figure 5-36: Token bucket algorithm in enb The tokens accumulate in a bucket, and the maximum allowed tokens are defined by the bucket size. The bucket size depends on the token rate and the bucket time. The tokens are consumed by the data packets and if the number of tokens in the bucket is positive the packet is transmitted. Each transmission involves a reduction in the number of tokens that amounts to the size of the transmitted packet. Such mechanism may create a "token debt" resulting in a negative number of tokens in the bucket. In this case, the transmission is delayed. Therefore, the data will be buffered on the UE. In this way, this feature will affect the application throughput of the UE trying to send at a higher rate than the token rate. Thus the average application throughput for the UE is limited by the Token Rate. 3.6 Delay Based Scheduling The new optional feature, Delay Based Scheduling and Grant Estimation, will increase the average data throughput of non VoIP bearers in both UL and DL for mobile broadband services in mixed scenario's, i.e. when both VoIP and mobile broadband (MBB) services are used Feature Functionality Different services have different needs and requirements, the two classical services expected to be supported in LTE are listed here for illustration of their main QoS demand: Ftp service: Delay insensitive/high throughput desired Voice service: Delay sensitive/less need for high throughput LZT R1A Ericsson AB

184 LTE L13 Radio Network Functionality The flexibility of the system allows the enodeb scheduler in Ericsson to offer an expandable set of strategies or traffic / service prioritization to be selected as seen below: Resource fair Proportional fair Degree of fairness configurable Delay based The service traffic flow prioritization in the scheduler is configurable per radio bearer using the enodeb QoS interface settings. Different services have different needs Ftp: Delay insensitive/high throughput desired Voice: Delay sensitive/less need for high throughput The scheduler offers an expandable set of strategies Resource fair Proportional fair Degree of fairness configurable Delay based Flow prioritization is configurable per radio bearer using the QoS interface Figure 5-37: Delay Based Scheduling - Flow Prioritization Background problem If we assume that the VoIP service quality in LTE will depend mainly on the mouth-to-ear delay. So, it has to be set in the system, a target delay up to a limit. This target delay will be compared against the current system delay in order to determine if the VoIP service considered still can be acceptable in case of low system performance. In case of shared air interface resources as LTE is by nature, the buffer status reporting as described in 3GPP is not efficient for VoIP. By setting the enodeb scheduling algorithm considering VoIP delay requirement characteristics, it can be utilized to get efficient scheduling which is important since VoIP is a resource demanding service: Periodicity between VoIP packets when the user is talking or silent. Size of the VoIP packets when the user is talking or silent Ericsson AB 2012 LZT R1A

185 Capacity Management Mobile broadband (MBB) throughput will suffer from a high VoIP load if VoIP is always prioritized. The enodeb scheduler will consider an internal weighting function to determine traffic resource in the LTE radio resource shared environment to decide what subscriber traffic to be prioritized over the air interface in the next Time Transmission Interval (TTI). The weight is calculated as a function of the age of the data, which is suitable for VoIP, since the quality depends on the delay. In this case the main advantage of a scheduling prioritizing traffic based on the weight are: In case segmentation is necessary the segments will be scheduled in consecutive TTI s. Young VoIP packets can be delayed and then bundling of packets is possible, allowing for scheduling of MBB. If packets can be delayed several packets can be scheduled at once (bundling of VoIP packets). The weight is calculated as a function of the age of the data. Suitable for VoIP since the quality depends on the delay. Advantages: In case segmentation is necessary the segments will be scheduled in consecutive TTI s. Young VoIP packets can be delayed and then bundling of packets is possible, allowing for scheduling of MBB. If packets can be delayed several packets can be scheduled at once (bundling of VoIP packets). Figure 5-38: Delay-based scheduling (DBS) Regarding to this delay based weighting function used internally by the enodeb scheduler: The DBS weight function depends on the age of the packet. MBB weights are lower than DBS weights after the bundling time. Prior to the bundling time the MBB weights will compete with the DBS weights. The diagram below illustrates DBS weight function. LZT R1A Ericsson AB

186 LTE L13 Radio Network Functionality Scheduling weight Larger Weight after bundling Time, but not guaranteed scheduling The bundling Time is the time to give weight bonus Giving a Small weight before reaching the bundling, but not blocking Default Priority Bundling Time Age of oldest packet At low load a VoIP user can be scheduled even before the bundling time At high load a VoIP user might not be able to be scheduled even after the bundling time Figure 5-39: DBS weight function In a radio environment condition, the performance of the Radio Blocks when facing the existent RF conditions can be of great impact on quality noticed by VoIP subscriber. Also, there may be many other on going services in parallel to VoIP also demanding for system shared resources. Channel quality is taken into account by the delay based scheduler to avoid that VoIP users that require extensive segmentation and retransmissions starve the system consuming all shared resources because of delay sensitivity. Channel quality is taken into account by the delay based scheduler to avoid that VoIP users that require extensive segmentation and retransmissions starve the system. Figure 5-40: Channel Quality factor Decisions on downlink scheduling or uplink access grant are made independently but both are made by the enodeb. If we consider that now scheduling have to consider the packet age when making scheduling decisions, the scheduling for downlink is expected to be easy since the DL RLC and DL Scheduler are in the enodeb. On the other hand, for uplink direction, it is less straight forward since the UL RLC is in the UE and the UL Scheduler is in the enodeb. Downlink Easy since the DL RLC and DL Scheduler are in the enb. Uplink Less straight forward since the UL RLC is in the UE and the UL Scheduler is in the enb. Figure 5-41: Determining the packet age Ericsson AB 2012 LZT R1A

187 Capacity Management For the uplink special case, the UL DBS includes two main tasks: Estimate the age of UL VoIP packets. Estimate the VoIP buffer size in the UE. o Even if the UE has not been scheduled for > 20 ms. Two tasks Estimate the age of UL VoIP packets. Estimate the VoIP buffer size in the UE. Even if the UE has not been scheduled for > 20 ms. Figure 5-42: Buffer estimation for UL DBS Benefits The benefit is most noticeable for DL MBB and at high VoIP loads. The VoIP capacity will be marginally increased. There will be more efficient VoIP scheduling due to the VoIP specific buffer estimation and bundling of VoIP packets. UEs that rely on segmentation (bad radio conditions) will get a shorter delay due to scheduling of the segments in consecutive TTI s. Increased throughput from MBB users at high VoIP load: VoIP packets from/to UEs in good radio conditions can be bundled without compromising VoIP quality. When VoIP packets are bundled resources are freed that can be used for scheduling of MBB services. This will be more pronounced if the feature is combined with robust header compression. It will also provide VoIP performance measurements by means of new counters and the introduction of a VoIP KPI. The KPI keeps track of the fraction of the UEs using VoIP in the cell that are satisfied with their VoIP performance. It uses the 3GPP definition for VoIP satisfaction and VoIP capacity. LZT R1A Ericsson AB

188 LTE L13 Radio Network Functionality Increased VoIP capacity More efficient VoIP scheduling due to The VoIP specific buffer estimation Bundling of VoIP packets UEs that rely on segmentation (bad radio conditions) will get a shorter delay due to scheduling of the segments in consecutive TTI s. Increased throughput from MBB users at high VoIP load. VoIP packets from/to UEs in good radio conditions can be bundled without compromising VoIP quality. When VoIP packets are bundled resources are freed that can be used for scheduling of MBB services. Figure 5-43: Benefits of the Feature In the following example we observe the enodeb Round Robin Scheduler strategy with QoS aware scheduling. In this example, VoIP users will always have higher priority than Data Users. The VoIP service quality gain is due to the Delay based Scheduling and Grant Estimation being used. RR: Round Robin Scheduler with QoS aware scheduling, VoIP users will always have higher priority than Data Users. DBS: Delay based Scheduling and Grant Estimation used. Delay based scheduling gains seen in at high VoIP load, where the scheduling capacity is limited Figure 5-44: Simulation results - UL Data ThroughPut Ericsson AB 2012 LZT R1A

189 Capacity Management Figure 5-45: simulation results - DL Throughput Implementation wise, the DBS implementation in the enodeb is independent of the UE and every other network element. Hence no specific behavior is required from the UE for DBS Configuration This feature has impact on node configuration. Ericsson introduces a new parameter pdb (packet delay budget), this is a MOM parameter. It determines how much delay it will be allowed by DBS before VoIP users get a large scheduling weight, which determines how many VoIP packets will be scheduled together (bundled). This parameter also determines when a VoIP packet is considered in time or delayed. The second parameter servicetype is a MOM parameter indicating which service the bearer carries. It must be set to VoIP for delay-based scheduling. LZT R1A Ericsson AB

190 LTE L13 Radio Network Functionality pdb (packet delay budget) is a MOM parameter. Determines how large delay will be allowed by DBS before VoIP users will get a large scheduling weight which determines how many VoIP packets will be scheduled together (bundled). Determines when a VoIP packet is considered in time or delayed. servicetype is a MOM parameter indicating which service the bearer carries. Must be set to VoIP for delay-based scheduling. Figure 5-46: Configuration MOM parameters The administration of this feature is done by activating the Delay-based scheduling installed license, by setting the correct QoS configuration, by enabling to VoIP bearer: ServiceType = VOIP, setting the pdb parameter (default = 80 ms according to 3GPP TS ), configuring the parameter SchedulingAlgorithm = DELAY_BASED, and in the QoS enodeb setting, by putting the VoIP bearer in a dedicated LCG It is possible to measure the VoIP performance by using the counters delivered by VoIP/video service counters and KPI s Ericsson AB 2012 LZT R1A

191 Capacity Management 3.7 Semi-Persistent Scheduling A VoIP packet is generated every 20 ms during TALK period. With Delay Based Scheduling/Service Aware Buffer Estimation (DBS/SABE) the VoIP packets are bundled and scheduled typically every 40 ms. The UE will send Scheduling Request (SR) and stay awake until scheduled. A VoIP talk packet is generated every 20 ms With DBS/SABE the VoIP packets are bundled Schedule interval ~40 ms The UE will send an SR and then stay awake until scheduled UE wakes up DBS/SABE triggered grants WAS TE WAS TE UE buffer gets a new voice packet and trigger SR UE buffer gets a new voice packet without sending an SR UE buffer gets a new voice packet and trigger SR UE buffer gets a new voice packet without sending an SR Figure 5-47: Background/Problem Description When silence is detected by SABE, the VoIP bearer enters SID (Silence Insertion Description) period. With Semi Persistent Scheduling (SPS), when SID to TALK transition is detected, an SPS Grant (using SPS-CRNTI) is sent to activate SPS. SPS will prevent the UE from sending SR for VoIP. SPS is reactivated when the SPS period is about to expire (after 75% of semipersistentschedulingintervalul). LZT R1A Ericsson AB

192 LTE L13 Radio Network Functionality The diagram below illustrates the Battery saving function. A VoIP bearer could be in SID or in TALK When SID to TALK transition is detected SPS is activated by sending an SPS grant Activating the SPS will prevent the UE from sending SR for VoIP SPS is reactivated when the SPS is about to expire Normal grants are sent to schedule the UE When TALK to SID transition is detected SPS is released by sending an explicit release grant. SABE detects VoIP talk initiates SPS-grant VoIP talk WAS TE WAS TE UE buffer gets a new voice packet and trigger SR UE buffer gets a new voice packet, no SR sent UE buffer gets a new voice packet and trigger SR UE buffer gets a new voice packet, no SR sent Figure 5-48: Battery Saving for DBS/SABE using Semi Persistent Scheduling (SPS) The UE saves battery with SPS, because it does not send SR and can sleep for longer periods. However, normal grants are used to schedule the UE, typically every 40 ms. When TALK to SID transition is detected, the SPS is released with a release grant (deactivation of SPS). DBS/SABE will continue to schedule the UE every 40 ms, until a SR/BSR for VoIP is received, which confirms that the UE has heard the deactivation. Feature name Battery Saving for DBS/SABE Ericsson AB 2012 LZT R1A

193 Capacity Management 3.8 TTI Bundling Due to the short TTI and relatively low UE tx power, the LTE UL coverage may sometimes suffer. This may lead to RLC segmentation of e.g. VoIP packets which in turn leads to increased overhead size. This problem can be mitigated by using TTI bundling. TTI Bundling aims for allowing VoIP calls in areas with bad coverage when VoIP calls would not be possible without TTI Bundling. With TTI bundling, the UE transmits same information bits, but with different redundancy versions (RV), in 4 consecutive TTIs instead of in only one TTI. The enb will decode and send HARQ ACK/NACK based on this whole bundle. This allows for higher energy per bit and thus increased coverage, lower RLC and MAC overhead and reduced signaling (only one UL grant and one HARQ ACK/NACK). In TTI Bundling, the UE transmits in 4 consecutive TTIs allows higher energy per bit Same information bits different redundancy version (RV) Compare to normal operation where one uplink grant corresponds to transmission in only one TTI (1) enodeb sends grant for uplink transmission (3) enodeb receives and performs decoding based on all 4 transmissions (4) enodeb sends ACK or NACK based on whole bundle RV 0 RV 1 RV 2 RV 3 TTI = 1 ms (2) UE transmits same data in 4 TTIs with different redundancy version time Figure 5-49: TTI Bundling - Feature Overview LZT R1A Ericsson AB

194 LTE L13 Radio Network Functionality The HARQ retransmission time between bundles is 16 ms. The overhead resulting from RLC and MAC headers as well as CRC is less when TTI Bundling is used compared to the case where RLC segmentation is used. The TTI Bundling feature is only used for UEs that are using a VoIP service. Higher energy per bit enables increased transport block (TB) size Reduced segmentation Lower overhead headers added per RLC PDU and MAC transport block Overhead size as much as 40% headers in normal operation Overhead decreased to 14% in TTI bundling mode Reduced number of control signaling messages Reduced vulnerability to control signaling errors in TTI Bundling mode Segmentation into 4 RLC PDUs: i.e. One RLC SDU using 4 RLC PDUs - Four PDCCH allocations. -Four HARQ feedbacks. Normal operation RLC PDU TB RLC SDU RLC PDU TB RLC PDU TB Overhead RLC PDU TB TTI Bundling RLC SDU RLC PDU TB Over head One RLC SDU transmitted in one RLC PDU: -One PDCCH allocation. -One HARQ feedback. time time Figure 5-50: TTI Bundling - Feature Overview cont d Switching between TTI bundling and normal operation is based on channel quality. The UE must support TTI bundling in order to use it. In accordance with the 3GPP specifications the feature TTI Bundling provides a maximum UL throughput of 126 kbit/s on the air interface. This means that UEs that are configured for TTI Bundling will have limited throughput. The TTI Bundling feature can affect DRX: the DRX parameter OnDurationTimer will be set to 4 for UEs that are configured to use TTI Bundling if the value of OnDurationTimer is less than 4 in the MOM. 3.9 DL Frequency Selective Scheduling In wideband transmission, as in LTE technology, the radio link quality varies not only over time, but also within the frequencies of the transmission bandwidth. By selectively scheduling the active UEs in different sub-bands where their channel qualities are better, both cell capacity and coverage is improved. Downlink Frequency Selective Scheduling uses measurements of the radio channel conditions (such as sub-band Channel Quality Indicators) to determine which part of the radio resource that should be used for a transmission Ericsson AB 2012 LZT R1A

195 Gain-to-Interference ratio [db] Capacity Management The Frequency Selective Scheduling improves cell-edge bitrate and cell capacity and increases cell capacity when multiple users share the same TTI. Gains are highest for high load low UE speed and frequency diverse channel. Radio link quality varies within the frequencies of the transmission bandwidth. Selective scheduling of UEs in different sub-bands where their channel qualities are better Both cell capacity and coverage can be improved Improved cell-edge bitrate and cell capacity Frequency [RBs] Time [ms] Figure 5-51: DL Frequency-Selective Scheduling Sub-band aperiodic PUSCH CQI reporting is used to support DL Frequency Selective Scheduling (FSS). The CQI reports contain both wideband CQI and sub-band values. The sub-band values are offsets to the wideband value (which is averaged over the whole system bandwidth). The aperiodic CQI report is requested when the UE has DL data. Sub-band CQI value for each codeword are encoded differentially with respect to their wideband CQI using 2-bits Sub-band differential CQI offset level = sub-band CQI index wideband CQI index Sub-band differential CQI value Offset level 0 (00) 0 1 (01) 1 2 (10) 2 3 (11) -1 Mapping sub-band differential CQI value to offset level Figure 5-52: Sub-band CQI reporting The overall principle of DL FSS feature is illustrated in Figure LZT R1A Ericsson AB

196 UE SE/PQ Scheduling Weight Priority LTE L13 Radio Network Functionality The illustration shows the scheduling and resource allocation behaviours for three UEs Scheduling starts from the highest to the lowest priority UEs. Resource allocation is confined to the UE s sub-bands where their channel qualities are better. In the example, as sub-bands S 5 and S 6 have been allocated to UE 2, UE 3 can only get allocation for subbands S 0 and S 2. User 3 User 1 User 2 Allocation Allocation Allocation S 0 S 1 S 2 S 3 S 4 S 5 S 6 User 1 User 2 User 3 Frequency [Hz] UE sub-band priority lists UE 1 UE 2 UE 3 S 3 S 6 S 5 S 4 S 5 S 6 S 1 S 2 S 0 Figure 5-53: DL Frequency Selective Scheduling - Feature Overview This feature requires QoS Aware Scheduling and Downlink Frequency Selective Scheduling licenses. The parameter dlresourceallocationstrategy can be set to Resource Fair (non- FSS) or Frequency Selective (FSS). FSS can be configured in conjunction with all existing Scheduling Algorithms Ericsson AB 2012 LZT R1A

197 Capacity Management 3.10 Relative Priority Scheduling This feature was first introduced in L12B. The feature depends: FAJ , QoS-Aware Scheduler FAJ , Minimum Rate Proportional Fair Scheduling Relevant features are needed in EPC in order to set the QCI parameter different for different bearers Feature Background The Proportional Fair Scheduler (PFS) allows the operator to balance fairness and cell throughput in system radio performance. PFS consists of two components: the rate component and the channel component. The channel component provides higher cell throughput by scheduling the UEs with better RF. The Proportional Fair Scheduler (PFS) allows us to balance fairness and cell throughput. PFS consists of two components: the rate component and the channel component. The channel component provides higher cell throughput by scheduling the UEs with better RF. The rate component provides fairness by providing weights that reflect the average bit rates of services. When the scheduling decision uses only this component, all services get similar bit rates independant of channel conditions. Figure 5-54: Relative Priority Scheduling -Background Proportional Fair Scheduling The rate component provides fairness by providing weights that reflect the average bit rates of services. When the scheduling decision uses only this component, all services get similar bit rates independent of channel conditions. LZT R1A Ericsson AB

198 LTE L13 Radio Network Functionality PFS supports 5 different fairness levels: Equal Rate: e.g. 100% rate component. Proportional Fair High: e.g. 80% rate component, 20% channel. Proportional Fair Medium: e.g. 60% rate component, 40% channel. Proportional Fair Low: e.g. 40% rate component, 60% channel. Max C/I: e.g. 100% channel component. Minimum bit rate: Configures the minimum rate in kbps per QCI. Used to prevent starvation. Figure 5-55: Background Proportional Fair Scheduling and minimum rate Feature Overview Proportional Fair Scheduling and RPS: RPS refines the rate component of a specific PFS fairness level. RPS is used to differentiate services within a PFS fairness level to distribute the bit rates per service based on the configured relative priority. QCI Table (QciProfilePredefined/QciProfileOperatorDefined) QCI Relative priority Sch.Alg. priority APO Min Rate 7 50% EqualRate n/a false % EqualRate n/a false % EqualRate n/a false 0 Weight Weight w Expected average weight QCIs 7,8,9 QCI 7 QCI 8 QCI 9 Rate r 9 r 8 r 7 Expected average bit rates Rate Figure 5-56: Example 1 Proportional Fair Scheduling with RPS Ericsson AB 2012 LZT R1A

199 Capacity Management Proportional Fair Scheduling, minimum rate, and RPS: RPS can also be combined with minimum rate. Services that are below their minimum rate are differentiated by RPS. In the case when two or more services are under their minimum rate, the service with the higher relative priority receives a higher scheduling weight. QCI Table (QciProfilePredefined/QciProfileOperatorDefined) QCI Relative priority Sch.Alg. priority APO Min Rate 7 50% EqualRate n/a false minrate 8 33% EqualRate n/a false minrate 9 17% EqualRate n/a false minrate Weight Weight Weight w + QCI 7 = MinRate Rate QCI 8 QCI 9 Rate QCI 7 QCI 8 QCI 9 MinRate Rate Figure 5-57: Example 2 Proportional Fair Scheduling with RPS and Minimum Rate LZT R1A Ericsson AB

200 mr_1 mr_2 mr_3 mr_1 mr_2 mr_3 LTE L13 Radio Network Functionality QCI Table (QciProfilePredefined/QciProfileOperatorDefined) QCI Relative priority Sch.Alg. priority APO Min Rate 7 50% EqualRate n/a false mr_1 8 33% EqualRate n/a false mr_2 9 17% EqualRate n/a false mr_3 Weight Weight Weight w + QCI 7 QCI 8 = QCI 9 Rate lower rate upper rate Rate QCI 7 QCI 8 QCI 9 Rate Figure 5-58: Example 3 Proportional Fair Scheduling with RPS and Minimum Rate Benefits and limitation of the feature RPS differentiates services that use PFS equal rate. For PFS fairness levels other then equal rate, RPS differentiates services of similar channel conditions. For services using minimum rate, RPS differentiates services that are below minimum rate. Absolute priority override (APO) must be enabled for all or none of the QCI entries that belong to the same RPS group. The relative priority is meaningful only within an RPS group. In order to enable the feature, the following licenses are required: QoS Aware Scheduling Proportional Fair Scheduling Relative Priority Scheduling. The relativepriority parameter is configured for a QCI entry enabling the feature Ericsson AB 2012 LZT R1A

201 Capacity Management Licenses required: QoS Aware Scheduling Proportional Fair Scheduling Relative Priority Scheduling. The relativepriority parameter is configured for a QCI entry enabling the feature. Figure 5-59: Configuration Statistics are supported for this Feature for downlink in L12B. PFS has relation with other Ericsson features: Proportional Fair Scheduling RPS modifies the rate component of PFS. Minimum Rate RPS can be combined with minimum rate as previously mentioned. Figure 5-60: Relation to other Features LZT R1A Ericsson AB

202 LTE L13 Radio Network Functionality 4 Service Specific DRX This feature was first introduced in L12B. This feature is dependant on: FAJ , Efficient DRX/DTX for Connected UE Feature Description Service-specific DRX allows the operator to configure the following: Default connected-mode DRX parameters Service-specific connected mode DRX parameters Prior to this release, default connected mode nor service specific DRX parameters were not configurable by the operator. DRX (Discontinuous Reception): Connected UEs monitor PDCCH only at part of a pre-defined DRX cycle Wake OnDurationTimer drxinactivitytimer sleep longdrxcycle Configuration parameters: ondurationtimer; longdrxcycle; drxinactivitytimer; shordrxtcycle; drxshortcycletimer; drxretransmissiontimer This results in reduced UE battery power consumptions This could cause data transmission delays Figure 5-61: DRX Background Service specific connected mode DRX parameters refer to DRX settings that change depending on the services that are established. This is based on the QCI values of the bearers that have been established for the UE Ericsson AB 2012 LZT R1A

203 Capacity Management Feature Description The connected mode DRX parameters can be adjusted based on the services that are currently in use by the UE. Data transmission characteristics: Web browsing/video streaming: big bursts followed by long rest VoIP: small packets every 20/40 ms or 160ms DRX without this feature: Same DRX settings are used for all UEs connected to an enb, regardless their data services. The current default DRX configuration favors general best effort data transmission that has a long longdrxcycle and a long drxinactivitytimer configured. Hence, VoIP or Smartphone UEs do not get much battery saving with this configuration. Set by system constants, cannot be tuned by operators. Figure 5-62: Problem Description For example, it is known that the default mobile broadband connected mode DRX settings are less than optimal for the case where a VoLTE call is active. This feature allows a mobile broadband connected mode DRX configuration to be used by default, but to switch to a VoLTE-specific DRX value when a VoLTE call is established. The feature includes: An extended drxprofiletable with 19 DRX profiles (operator configurable) Each DRX profile contains a set of DRX settings QCI dependent DRX setting (operator configurable) Each QCI associates one DRX profile based on its services characteristics drxprofiletable is enb based. DRX reconfigurations May occur during adding, releasing or modifying an E-RAB. ANR CGI measurement activities Performance Monitoring Each DRX setting has DRX event based counters are extended to DRX configure index. Figure 5-63: Service Specific DRX LZT R1A Ericsson AB

204 LTE L13 Radio Network Functionality Default connected mode DRX parameters can now be configured by the operator. drxprofileid 0 (default) ANR (not configured by operator) drxprofileid 1 drxprofileid 18 drxconfi gidx ondurationtimer onduratio ntimer onduratio ntimer onduratio ntimer ondurationtimer ondurationtimer drxinactivitytimer drxinactivi tytimer drxinactivi tytimer drxinactivi tytimer drxinactivitytimer drxinactivitytimer longdrxcy cleonly longdrxc ycle longdrxcy cleonly longdrxcy cleonly longdrxcy cleonly longdrxc ycleonly longdrxc ycle longdrxc ycleonly longdrxc ycle N/A shordrxt Cycle N/A N/A N/A N/A shordrxt Cycle N/A shordrxt Cycle N/A drxshort CycleTim er drxretransmission Timer N/A N/A N/A N/A drxshort CycleTim er N/A N/A N/A drxretransmission Timer N/A drxshort CycleTim er drxretransmission Timer Configurable Not configurable Configurable Figure 5-64: DRX Profile table Service-specific connected mode DRX parameters can be defined SRB Predefined QCI Operator defined QCI btindex SRB SRB1 SRB 2 N/A N/A N/A N/A... N/A QCI N/A N/A QCI Default (0) QCI = 1..9 QCI = QCI = rlcmode N/A N/A AM From configuration, MOC QciProfilePredefined Op defined Op defined rlcsnlength N/A N/A QciProfPred. rlcsnlength QciProfPred. rlcsnlength Op defined Op defined... pdcpsnlength N/A N/A QciProfPred. pdcpsnlength QciProfPred. pdcpsnlength Op defined Op defined... drxprofileid N/A N/A 0 Op defined Op defined Op defined... drxpriority N/A N/A 0(lowest) Op defined Op defined Op defined... Follows DRX setting for DRB Figure 5-65: DRX profile mapping to QCI Instead of one for all setting, a set of DRX settings can be configured by operator, providing the best fit DRX settings for certain services. DRX configuration is chosen based on current UE s primary service configured by operator Ericsson AB 2012 LZT R1A

205 Capacity Management By applying best fit DRX configuration based on current services, UE battery life can be optimized without sacrificing system performance. This is supported through the introduction of DRX profiles, and the mapping of QCIs to DRX profiles. According to 3GPP, UE reports periodic CSI through PUCCH can be only done during UE awake. Periodic CSI report period is 40/80 ms if configured. The DRX long cycle configuration should be configured as either a multiple of the period of UE periodic CSI report configuration, or period of UE periodic CSI report is a multiple of long DRX cycle. The rows for long DRX cycles with in bold color are not recommended. sf10 sf20 sf32 sf40 sf64 sf80 sf128 sf160 sf256 sf320 sf512 sf640 sf1024 sf1280 sf2048 INTEGER(0..9), INTEGER(0..19), INTEGER(0..31), INTEGER(0..39), INTEGER(0..63), INTEGER(0..79), INTEGER(0..127), INTEGER(0..159), INTEGER(0..255), INTEGER(0..319), INTEGER(0..511), INTEGER(0..639), INTEGER( ), INTEGER( ), INTEGER( ), Figure 5-66: Limitations Each QCI is also associated with a DRX priority. The enb selects the DRX profile associated with the QCI having the highest DRX priority. LZT R1A Ericsson AB

206 LTE L13 Radio Network Functionality DRX and ServiceSpecificDrx are optional features with both licenses, Operator is able to Configure an extended DRX profiles table Map a DRX profile to QCI and set DRX priority with DRX license only, Operator is able to Configure the default DRX profile This DRX profile is used for all QCIs if neither license DRX configuration is disabled. Feature is supported on both FDD and TDD systems. Figure 5-67: Administration DRX and ServiceSpecificDrx are optional features with both licenses. Operator is able to configure an extended DRX profiles table and map a DRX profile to QCI and set DRX priority. With DRX license only, Operator is able to Configure the default DRX profile. This DRX profile is used for all QCIs if neither license DRX configuration is disabled. The feature is supported on both FDD and TDD systems. Up to 19 DRX profiles can be configured by operator Each profile contains two DRX configurations for UEs supporting or not supporting short DRX respectively Under the same DRX profile, longdrxcycleonly does not have to be the same as longdrxcycle, so advantage of UE s capability on supporting shortdrxcycle can be well taken drxprofileid ondurationtimer drxinactivitytimer longdrxc longdrxc ycleonly ycle N/A shordrxt Cycle N/A drxshort CycleTim er drxretransmission Timer Figure 5-68: DRX Profiles Three ANR configurations (Eutran, Utran and Geran cases) are not configured by operators Ericsson AB 2012 LZT R1A

207 Capacity Management With DRX license only, operator can only access drxprofile0 (default) for DRX configuration Figure 5-69: QCI dependent DRX configurations Each QCI associates one DRX profile based on its services characteristics. The same DRX profiles can be set to multiple QCIs. Each QCI has drxpriority, a higher value indicates a higher priority for the associated DRX profile if multiple QoS are serving a UE. Among QCIs, drxpriority should be set with a unique value unless the same DRX Profile is associated for those QCIs. If two QCIs are configured with the same drxpriority values but different drxprofiles, enb could randomly pick one of the profiles for DRX settings. DRX configuration for ANR CGI measurement takes a higher priority than QCI based DRX configuration. drxprofileid 0 ANR drxprofileid 1 drxprofileid 18 (default) drxconfigid x 1 ondurationtimer onduratio ntimer ondurati ontimer ondurati ontimer ondurationtimer ondurationtimer drxinactivitytimer drxinactivit ytimer drxinacti vitytimer drxinacti vitytimer drxinactivitytimer drxinactivitytimer longdrxcycle longdrxcycl e longdrxcy cle longdrxc ycle longdrxc ycle longdrxcycle longdrxcycle longdrxcycle longdrxcycle N/A shordrxtcy cle N/A N/A N/A N/A shordrxtcycl e N/A shordrxtcycle N/A drxshortcy cletimer N/A N/A N/A N/A drxshortcycl etimer N/A drxshortcycleti mer drxretransmission Timer N/A N/A N/A drxretransmission Timer drxretransmission Timer Cell 0 Cell 1... Cell N enb UE UE UE UE UE Figure 5-70: Apply DRX configurations LZT R1A Ericsson AB

208 LTE L13 Radio Network Functionality DRX Profile is enb based. The Cells follow the same table for DRX configuration. For each UE, DRX profile is selected based on current services used for this UE also drxconfigidx under the selected DRX profile is chosen based on UE s DRX capability and DRX configuration under drxconfigidx is communicated with UE through RRC reconfiguration message. Updated DRX profiles take effect by unlocking the first cell in the enb after the update. It is recommended that all cells are locked for DRX Profile reconfiguration but not forced for this action. In case a cell is not locked for DRX profile update, UE could be lost due to inconsistent DRX configuration. Once UE is re-attached, new DRX configuration is applied. UE stays connected if DRX it used doesn t change. DRX configuration selection is triggered by 1. add/release/modify E-RABs - DRX profile associated to one of the E-RABs that has the highest DRX priority is chosen; - Based on UE s capability on supporting short cycle DRX or not, the associated drxconfigidx is chosen. - Updated DRX configuration is communicated to UE in RRC reconfiguration message along with e-rab update information. 2. ANR CGI measurements Start: Use DRX configuration for ANR case Stop: Follow case 1 for drxconfigidx selection New DRX configuration takes effect at enb upon receiving UE s RRC reconfiguration complete message. Figure 5-71: DRX configuration selection Ericsson AB 2012 LZT R1A

209 Capacity Management drxprofileid=1 drxprofileid=2 Bearer 1 drxpriority =0 drxprofileid=0 Bearer 2 drxpriority =1 drxprofileid=1 Bearer 1 drxpriority =0 drxprofileid=0 Bearer 2 drxpriority =1 drxprofileid=1 Bearer 3 drxpriority =2 drxprofileid=2 Bearer 3 Add an E-RAB drxpriority =2 drxprofileid=2 UE supports short DRX cycle Before: Bearers 1 & 2, highest drxpriority = 1 After: Bearers 1, 2 & 3, highest drxpriority = 2 Figure 5-72: DRX configuration selection (add e-rab) drxprofileid=2 drxprofileid=2 Bearer 1 Bearer 3 Bearer 1 Bearer 3 drxpriority =0 drxprofileid=1 Bearer 2 drxpriority =1 drxpriority =2 drxprofileid=2 drxpriority =0 drxprofileid=1 drxpriority =2 drxprofileid=2 drxprofileid=1 Release an E-RAB UE supports short DRX cycle Before: Bearers 1, 2 & 3, highest drxpriority = 2 After: Bearers 1 & 3, highest drxpriority = 2 Figure 5-73: DRX configuration selection (remove e-rab) The Feature has dependency on other features both already existent from previous releases and features new in L12B. The picture below illustrates this scenario. This feature requires: Pre-L12B DRX DRX for ANR CGI Measurements L12B DRX for TDD Figure 5-74: Relation to other Features LZT R1A Ericsson AB

210 LTE L13 Radio Network Functionality A mechanism for which different QCIs are allowed to have different DRX settings which are defined by the operator. The main Benefits of this features are: Instead of one for all setting, a set of DRX settings can be configured by operator, providing the best fit DRX settings for certain services. DRX configuration is chosen based on current UE s primary service configured by operator. By applying best fit DRX configuration based on current services, UE battery life can be optimized without sacrificing system performance. 4.2 Connection Setup The connection setup traffic case is shown in Figure Note that after a shared channel transmission, HARQ retransmissions may occur due to negative HARQ acknowledgements. All transmissions in the figure, including the possible asynchronous HARQ retransmissions in DL, are controlled by the Scheduler, except the RACH preambles and MIB on PBCH. Another exception is the possible HARQ retransmissions in UL, which are synchronous and do not require a scheduling grant. enodeb MME RRC_CONNECTED Random Access RRC Cell Selection MAC MAC BCCH: System Information PRACH: RACH preamble DL-SCH: RACH response RRC MAC MAC Admission Ctrl Allocation of SRB resources in BB RRC Connection Establishment Initial Context Setup * The IMSI is provided in the Attach Request ** enb UE S1AP id is included in all UE-related DL S1AP messages *** MME UE S1AP id is included in all UE-related UL S1AP messages except for Initial UE message RRC RRC RRC RRC RRC RRC RRC RRC LTE active UL-SCH: RRC Connection Request (Initial UE identity, Cause) DL-SCH: RRC Connection Setup (SRB1 parameters) UL-SCH: RRC Connection Setup Complete (Selected PLMN id, NAS: Attach Request *) DL-SCH: Security Mode Command (Security Configuration) UL-SCH: Security Mode Complete DL-SCH: RRC Connection Reconfiguration (Intra-frequency measurement configuration, Bearer Setup, NAS: Attach Accept) UL-SCH: RRC Conn Reconf Complete UL Inform Transfer (NAS: Attach Complete) RRC RRC RRC S1-AP S1-AP RRC RRC RRC RRC S1-AP RRC S1-AP RRC connected MME selection (based on S-TMSI) Initial UE Message (enb UE S1AP id **,NAS:Attach Request,TAI) Initial Context Setup Request (MME UE S1AP id ***, NAS: Attach Accept, Security, Bearer params, e.g. TEID) Allocation of payload bearer resources Initial Context Setup Response (Bearer params, e.g. TEID) Uplink NAS Transport (NAS: Attach Complete) S1-AP S1-AP S1-AP S1-AP LTE active Figure Connection Setup Ericsson AB 2012 LZT R1A

211 Capacity Management 4.3 RRM Related Measurements in LTE Measurements are the key to ensure appropriate execution of RRM functions. They are needed for single cell as well as for multi-cell RRM functions. Both enb and UE constantly executes measurements. Since E-UTRAN will support several RRM functions, different measurements will serve different purposes. In general, the measurements should provide a good estimate of the used and available radio resources, cell coverage, short and long term channel quality, cell load, service quality, time alignment during handovers, load on the transport network etc. In addition, measurements should provide good mobility support not only within E-UTRAN but also between E-UTRAN and other access networks including UTRAN and GERAN. For enb and UE measurements, the downlink and uplink reference symbols play a key role enb Measurements In E-UTRAN, certain types of measurements shall be performed internally in the enb and will not be exchanged between the enbs. These measurements do not need to be specified in the standard; rather they will be implementation dependent. On the other hand, measurements, which are to be exchanged between the enbs over the X2 interface are standardized. The enb measurements are described below. The positioning related measurements are not listed since they depend upon the exact positioning method used in E-UTRAN. Also, the current description does not explicitly take into account the impacts of multiple transmit and receive antennas on the measured quantities and measurement procedures (this issue is FFS). The following enb measurements are implementation specific and need not be specified in the standard: DL total Tx power: Transmitted carrier power measured over the entire cell transmission bandwidth. DL resource block Tx power: Transmitted carrier power measured over a resource block. DL total Tx power per antenna branch: Transmitted carrier power measured over the entire bandwidth per antenna branch. DL resource block Tx power per antenna branch: Transmitted carrier power measured over a resource block. DL total resource block usage: Ratio of downlink resource blocks used to total available downlink resource blocks (or simply the number of downlink resource blocks used). LZT R1A Ericsson AB

212 LTE L13 Radio Network Functionality UL total resource block usage: Ratio of uplink resource blocks used to total available uplink resource blocks (or simply the number of uplink resource blocks used). DL resource block activity: Ratio of scheduled time of downlink resource block to the measurement period. UL resource block activity: Ratio of scheduled time of uplink resource block to the measurement period. DL transport network loss rate: Packet loss rate of GTP-U (or frame) packets sent by the access gateway on S1 user plane. The measurement shall be done per traffic flow. The enb shall use the sequence numbers of GTP-U (or frame) packets to measure the downlink packet loss rate. UL transport network loss rate: Packet loss rate of GTP-U (or frame) packets sent by the enb on S1 user plane. The measurement shall be done per traffic flow. The access gateway shall use the sequence numbers of GTP-U (or frame) packets to measure the downlink packet loss rate. UL RTWP: Received total wideband power including noise measured over the entire cell transmission bandwidth at the enb. UL received resource block power: Total received power including noise measured over one resource block at the enb. UL SIR (per UE): Ratio of the received power of the reference signal transmitted by the UE to the total interference received by the enb over the UE occupied bandwidth. UL HARQ BLER: The block error ratio based on CRC check of each HARQ level transport block. Propagation delay: Estimated one way propagation delay measured during random access transmission. UE Tx time difference: Time difference between the reception of the UE transmitted signal and the reference symbol transmission time instant. The following enb measurements are specified in the standard: DL RS TX power: Downlink reference signal transmit power is determined for a considered cell as the linear average over the power contributions (in [W]) of the resource elements that carry cell-specific reference signals which are transmitted by the enb within its operating system bandwidth. Received Interference Power: Uplink received interference power, including the thermal noise within one physical resource block s bandwidth of 12 resource elements. The reported value shall contain a set of Received Interference Powers UL of physical resource blocks, n 0,..., 1. PRB N RB Ericsson AB 2012 LZT R1A

213 Capacity Management Thermal Noise Power: The uplink thermal noise power within the UL system UL bandwidth consisting of N RB resource blocks. It is defined as (N o x W), where N o denotes the white noise power spectral density on the uplink carrier frequency UL RB and W N N f denotes the UL system bandwidth. RB sc For inter-cell interference coordination purposes, it may be useful to measure the user plane load (for instance in terms of number of sent user plane packets/bits per second). The definition of such measurements and associated procedures are for further study UE Measurements The UE measurement quantities are described below. RSRP (Reference Signal Received Power): is determined for a considered cell as the linear average over the power contributions (in [W]) of the resource elements that carry cell-specific reference signals within the considered measurement frequency bandwidth. RSRQ (Reference Signal Received Quality): is defined as the ratio N RSRP/(E-UTRA carrier RSSI), where N is the number of RB s of the E- UTRA carrier RSSI measurement bandwidth. The measurements in the numerator and denominator shall be made over the same set of resource blocks. 5 Power Control Power control and power configuration reduces inter-cell interference and power consumption. This leads to higher cell capacity and the control of maximum data rate for a UE at cell edge. In addition, it maximizes battery life for the UE. Channel prediction provides the information needed for the other functions to make appropriate decisions, such as power control and link adaptation. Input to Channel Prediction comes from RBS measurements (for UL) and UE measurements in Channel Feedback Reports (CFRs for DL). Link adaptation selects the transport format to ensure the quality of service requirements are enforced while using resources efficiently. It is also essential for maximizing user throughput over the air interface. The resulting data rate also depends on scheduling of the Physical Resource Blocks (PRBs). Power Control is used to minimize the transmitted power and to compensate for channel fading. Its objective is to maximize capacity by optimizing the transmit power. Power control regulates the PSD (Power Spectral Density) of the transmitted signal and starts immediately when enough measurements are collected. LZT R1A Ericsson AB

214 LTE L13 Radio Network Functionality The following types of power control algorithms are employed in E-UTRAN: Common channel power control (PBCH, PCFICH, RS, SS) DL power setting (PDSCH, PDCCH, PHICH) PRACH preamble power ramping Open loop power control for initial access PUCCH power PUSCH power Open and closed loop power control UE RBS Figure Power Control in LTE 5.1 Open loop Power Control Open loop power control is used for: Regulating power for PRACH at initial access (Random Access) Regulating power for PUSCH and PUCCH as part of UL power control Random Access Random Access is part of the LTE Basic functions in the LTE Radio Access Network (RAN). The random access process obtains new or renewed access to the network and uplink timing synchronization. In the LTE network, the User Equipment (UE) employs the random access process to gain access to cells for the following reasons: Initial access to the network from the RRC_IDLE state. Regaining access to the network after a radio link failure. As part of the handover process to gain timing synchronization with a new cell. Prior to downlink data transfers in RRC_CONNECTED state when the UE is not time synchronized with the network Ericsson AB 2012 LZT R1A

215 Capacity Management Prior to uplink data transfers in RRC_CONNECTED state when the UE is not time synchronized with the network. For scheduling requests, when the UE has not been assigned a dedicated scheduling request resource on Physical Uplink Control Channel (PUCCH). The random access process allows multiple UEs to simultaneously gain access to a cell by using different random access preamble sequence codes. These codes are Zadoff Chu codes and have very good cross correlation properties. The codes are transmitted by UE on the uplink in specific Physical Random Access Channel (PRACH) subframes. Thanks to the good cross correlation properties, several UEs can be separated by the enb, even if they transmit in the same subframe. There are two different forms of the random access process, Contention Free Random Access (CFRA) and Contention Based Random Access (CBRA). The Contention Free Random Access (CFRA) is not used in the current release of LTE. The Contention Free Random Access (CFRA) process is initiated by the network and uses a dedicated random access preamble code allocated to the UE for a limited period. The CFRA process is used when the UE is known to the network. The CFRA process involves three phases and uses a preamble code dedicated to one UE to increase the probability of success of the random access process, leading to faster cell access. CBRA can also be initiated by the network The Contention Based Random Access (CBRA) process is initiated by the UE to gain access to the network. It involves the UE selecting a random access preamble code from a list of codes available for selection by all UE in the cell. CBRA requires additional signaling to resolve contention that may occur when multiple UE attempts to access the cell in the same PRACH subframe using the same preamble code. The application of CBRA and CFRA to different random access events is shown in the following table: Random Access Scenario CBRA CFRA Initial network access Access following a radio link failure Handover between cells Downlink data transfer requiring UE ynchronization Uplink data transfer requiring UE synchronization Uplink data transfer without dedicated scheduling request resource The CBRA process can be used for all random access application, however it is generally preferable to use the CFRA process for handover and synchronization as it allows for faster access times. LZT R1A Ericsson AB

216 LTE L13 Radio Network Functionality There are a number of stages involved in the random access process. Two of these stages are common to both the CFRA and the CBRA processes. Figure 5-77 shows a flow diagram of the random access process and identifies the stages used in the CFRA and CBRA processes. RA Preamble Assignment 0 1 Random Access Preamble Random Access Response 2 CFRA* Contention Free Random Access Process Steps CBRA Contention Based Random Access Process Steps 3 Scheduled Transmission (MSG3) HARQ Contention Resolution (MSG4) HARQ 4 *CFRA is not supported in L12 Figure 5-77: Random Access Procedure The CFRA process involves stages 0, 1 and 2. The CBRA process involves stages 1, 2, 3 and 4. The main difference between the CFRA and CBRA is that in CFRA the UE is assigned a dedicated preamble code sequence by the cell and no potential exists for another UE to attempt a random access using the same preamble code sequence. This makes the CFRA process faster as the contention resolution stages 3 and 4 can be avoided Random Access Process Steps This section describes the main process steps in the feature process flow CFRA PREAMBLE ASSIGNMENT The first step in the CFRA process is the assignment of a dedicated preamble code sequence to the UE. This is shown as step 0 in Figure Each LTE cell reserves a number of preamble codes for CFRA. Pre assignment of the preamble sequence by the network makes the random access process faster as there is no need for contention resolution. The preamble code required for the random access is sent to the UE by the serving cell Ericsson AB 2012 LZT R1A

217 Capacity Management RANDOM ACCESS PREAMBLE TRANSMISSION The random access preamble transmission and random access response steps are common to both CFRA and CBRA processes. They are shown as step 1 and step 2 respectively in Figure To optimize the Random access performance, the following parameter is used: - rachrootsequence: The first root sequence number for Random Access Channel (RACH) preamble generation. RACH root sequence is broadcasted as a part of system information distribution and used for preamble detection. RA Preamble Assignment Random Access Preamble Figure 5-78 RachRootSequence From 3GPP TS a relation between the cell size, described by the parameter cellrange, and the number of RACH root sequences needed for a cell has been derived rachrootsequence takes the value in the range of 0 to 837 and the default value is 386. It is recommended to set the parameter rachrootsequence to different values in neighboring cells to reduce the probability for false RACH detections. Figure 5-79 Maximum Cell Range The UE transmits random access preamble bursts on the PRACH uplink channel. The network provides information about the PRACH to UE in a system information message. This allows the UE to determine when the PRACH channels are scheduled and the preamble format and code sequence to use. The following information is included in the system information message: LZT R1A Ericsson AB

218 LTE L13 Radio Network Functionality The LTE basic system uses preamble format 0 as defined in Chapter of 3GPP TS This provides a maximum cell coverage range of approximately 14 km. The LTE basic system allocates one PRACH occurrence in each 10 msec radio frame. The random access transmission process uses open loop power control. The UE estimates the transmit power required to achieve a specified receive power at the RBS for the first random access burst. Power ramping is used for subsequent retransmission bursts if they are required. This process continues until the UE successfully receives a response from the RBS, or the maximum number of retransmission attempts is reached. The open loop power control process is detailed later in this chapter RANDOM ACCESS RESPONSE MESSAGE The random access response message is generated by the RBS and transmitted on the Physical Downlink Shared Channel (PDSCH) to a specific Random Access Radio Network Temporary Identity (RA-RNTI) address. More than one RA- RNTI address can be included in this message, allowing the setup of random access attempts from multiple UEs in the same PRACH sub frame. The UE monitors the PDCCH for its specific RA-RNTI address which matches the sub frame number in which the random access preamble burst was transmitted. The random access response message includes the following information: Random access preamble sequence code identifying the preamble sequence code which has been detected by the RBS Initial uplink schedule grant used for transmitting the subsequent data on the uplink channel Timing Alignment information required to ensure packet collisions do not occur during subsequent data transmissions A Temporary Cell Radio Network Temporary Identity (C- RNTI) CBRA CONTENTION RESOLUTION The contention resolution steps are used by the CBRA process in order to resolve a situation where two UEs have attempted a random access using the same preamble code sequence. They are shown as step 3 and step 4 respectively in Figure In step 3, the UE uses the first scheduled uplink transmission on Physical Uplink Shared Channel (PUSCH) following the successful receipt of a random access response message. The UE provides the network with a unique identifier in this message Ericsson AB 2012 LZT R1A

219 Capacity Management In step 4, the RBS echoes the UE identity provided in step 3. Only a terminal which finds a match between the identity received in step 4 and the identity transmitted as part of the step 3 declares the random access procedure successful. If the random access procedure is deemed to be successful the UE transmits an acknowledgement in the uplink. Terminals which do not find a match between the identity received in step 4, and the respective identity transmitted as part of step 3 are considered to have failed the random access procedure and must restart the random access procedure from step 1. Both step 3 and step 4 uses the Hybrid Automatic Repeat Request (HARQ) process. Further details on the contention resolution process and the HARQ process are detailed in Chapter 5.1 of 3GPP TS Open loop power control for Random Access The aim of initial random access open loop power control in the uplink is to ensure new connections are established causing minimum interference. The three basic steps employed are shown in Figure 5-80 below. PREAMBLE _ RECEIVED _ TARGET _ POWER preambleinitial ReceivedT arg etpower PREAMBLE _ TRANSMISSION _ COUNTER 1 powerrampi ngstep PREAMBLE 1) UE measures RS 2) Transmits at calculated power Connection established with minimum interference to other cells UE 3) The power is ramped up until a response is heard or maximum number of re-attempts is reached RBS Figure 5-80 Uplink Open Loop Power Control Open Loop Power Control is used in the uplink to minimize uplink interference when setting up a connection. PREAMBLE _ RECEIVED _ TARGET _ POWER preambleinitial ReceivedT arg etpower PREAMBLE _ TRANSMISSION _ COUNTER 1 powerrampi ngstep PREAMBLE In the early releases, the preambleinitialreceivedtargetpower, PREAMBLE and powerrampingstep are not configurable. LZT R1A Ericsson AB

220 LTE L13 Radio Network Functionality There are two types of the Random Access preamble groups. Depending on the message size and pathloss UE will select either Random Access Preamble group A or B. Once the group is selected the UE will randomly select a preamble within the group. pmaxservingcell powerrampingstep Data preambleinitial ReceivedPowerTarget Uplink (PRACH) Downlink (PDCCH)... preamble 0.8 ms subframe RA response window 1 ms RACH Preamble RACH Response... Indicates RA Response on PDSCH (Not UE specific) RA-RNTI... Downlink (PDSCH) No Response RA msg 3 RAPID Timing (UL timing) Scheduling Grant Figure 5-81 Preamble based power ramping Once preamble has been selected or it has been given by the network it should be sent using initial power setting. There are two preamble formats, 0 and 1. Using preamble format 1 makes it possible to increase the cell range to 100 km. In the current release, both preamble formats 0 and 1 will be supported. Refer to the document Maximum Cell Range. A new parameter is introduced in L12for the Random access process: rachrootsequence. rachrootsequence takes the value in the range of 0 to 837. The default value is 384. It is recommended to set the parameter rachrootsequence to different values in neighboring cells to reduce the probability for false RACH detections. The values must differ by at least 10 between any two neighbors (and at most 827 since there is a wrap-around between the first and last value in the value range). See also 3GPP TS Random Access parameters are broadcasted in SIB Ericsson AB 2012 LZT R1A

221 Capacity Management 5.3 DL Power Control Common Channel Power Control 3GPP has specified the modulation, coding and BLER requirements for PBCH, PCFICH, RS and SS so that they all have the same performance at the cell edge when they have the same PSD (Power Spectral Density) per Resource Element. The power levels for common channels are defined relative to a common Reference PSD; PSD_reference. This PSD_reference, is the PSD available out from the Radio Unit for each RE on each antenna if the totally available radio unit transmission power is split equally over all Resource Elements in the configured downlink bandwidth and all antennas configured for transmission in the cell. All common and shared downlink channels and signals, with one exception, will have their PSD set equal to PSD_reference. The exception is the RS PSD for the case of two antenna ports. In this case there will be an unused RE on one antenna port when there is an RS on the corresponding RE on the other antenna port. As a result, there is a 3 db boost for free, i.e. PSD RS will be 2 times PSD_reference. That is because energy can be moved within one power amplifier, between different REs in the same OFDM symbol Downlink Power Setting All PDSCH resource elements will use the same PSD level setting; PSD_reference. The PSD for PDCCH and PHICH are also set static to PSD_reference. DL common channels use the same PSD (set by PSD_reference) PBCH, PCFICH All physical signals use PSD_reference SS, RS (except when two antenna ports are used, which results in 3dB boost) PDSCH uses PSD_reference PDCCH uses PSD_reference PHICH uses PSD_reference Figure DL Power Control. LZT R1A Ericsson AB

222 LTE L13 Radio Network Functionality Power Spectral Density (PSD) is the power of a signal, divided by the bandwidth. P PSD BW PSD used here is normalized to one RB (180 khz). P P TX,RX BW=1RB f Figure 5-83 What is Power Spectral Density? 5.4 UL power control Uplink power control is used both on the PUSCH and on the PUCCH. In both cases, a parameterized open loop combined with a closed loop mechanism is used. Roughly, the open loop part is used to set a point of operation, around which the closed loop component operates. Different parameters (targets and 'partial compensation factors') for user and control plane are used. UE 2 Increase power reduce power Uplink PSD* target is maintained by open loop pzeronominalpucch pzeronominalpusch Closed loop fast enough to compensate for slow fading UE 1 RBS *PSD=Power Spectrum Density Figure 5-84: Uplink Power Control Ericsson AB 2012 LZT R1A

223 Capacity Management PUSCH Power Control In more detail, for the PUSCH the UE sets the output power P PUSCH according to the formula: P PUSCH = min {P CMAX,10logM + P 0_PUSCH + PL + f( )+ TF } [dbm] Where: P CMAX is the configured UE transmitted power defined in 3GPP TS M is the number of scheduled resource blocks, PL is the DL pathloss, estimated by the UE. The UE can use measurements of DL pathloss for estimation of UL pathloss. P 0_PUSCH is the target PSDrx, for each resource block, set according to the parameter pzeronominalpusch. It is common for all UEs in the cell. (enabling fractional pathloss compensation), set to 1, so the open loop compensates completely for the path loss. δ PUSCH is a UE specific offset or closed loop correction (the function f(i) may represent either absolute or accumulative offsets), included in the scheduling grant. It enforces the open loop PSDtarget and is updated fast enough to compensate for the slow fading. The f(i) is the closed loop power control part, while the rest is the open loop power control part. δ PUSCH is a UE specific power adjustment included in the uplink scheduling grant on PDCCH. TF is a Transport Format (TF) specific offset, set to 0. The closed loop correction δ PUSCH is sent in UL grants on the PDCCH, or in special TPC_PUSCH messages on the PDCCH. For all other parameters, cell specific values are broadcast. For a subset of the parameters (e.g. P 0 ), it is possible to send UE specific values via RRC. For persistent scheduled resources, it is possible to use an offset to P 0. The sounding reference signal power follows the PUSCH power with an enodeb controlled offset. Power control for PUSCH reduces the PSD TX for UEs close to the RBS. In this release, the target PSD RX is kept constant. The measured PSD RX is used directly to calculate power control commands. Power headroom reports are used by power control in order to know when UEs become power limited. PSD RX is also used to calculate SINR (based on measured N+I and PHR), which is used by PUSCH link adaptation. LZT R1A Ericsson AB

224 LTE L13 Radio Network Functionality P PUSCH = min{p CMAX,10 logm + P 0_PUSCH + PL + f(i) + TF } [dbm] Closed loop P CMAX : configured UE transmitted power defined in P 0 : Target PSD pzeronominalpusch M : Number of assigned resource blocks : Cell-specific factor TF : Transport format-depending compensation f(i) : Accumulation function or absolute function ( f(x) = x ) PUSCH : Power-control step ( PUSCH TPC command ). Input to f(i). PL : Estimated DL path loss PUSCH included in Uplink Scheduling Grant Figure 5-85: Uplink Power Control PUSCH The function f can either be cumulative or absolute. Only the accumulative function is enabled. In case of cumulative (accumulation enabled by higher layers): f(i) = f(i-1) + δ PUSCH (i-k PUSCH ) In case of absolute (accumulation disabled by higher layers): f(i) = δ PUSCH (i-k PUSCH ) where: i denotes the subframe. K PUSCH is 4 for FDD mode. The function f can either be cumulative or absolute. In case of a cumulative function (accumulation enabled by higher layers): f(i) = f(i-1) + δ PUSCH (i - K PUSCH ) In case of absolute function (accumulation disabled by higher layers): f(i) = δ PUSCH (i - K PUSCH ) where: i denotes the subframe. K PUSCH is 4 for FDD mode. Figure PUSCH power control. δ PUSCH is set according to: Ericsson AB 2012 LZT R1A

225 Capacity Management TPC command Accumulated δ PUSCH [db] Absolute δ PUSCH [db] Figure Delta PUSCH Setting. P0 for PUSCH and PUCCH statically signaled in SIB2 UL PL estimated from DL PL α = 1 (full pathloss compensation), Δ TF =0 f(i) compensates for slow fading, estimation errors, UL/DL pathloss difference etc f(i) = f(i-1) + δ PUSCH (i K PUSCH ) SI SIB2: α, P0 PUSCH PDCCH: M, δ P 0 = PSD RXtarget for PUSCH PUSCH P max, PL P PUSCH Figure 5-88: PUSCH Power Control P PUSCH = min {P CMAX,10logM + P0 PUSCH + α PL + f(i)+δ TF } [dbm] PUCCH Power Control PUCCH is power controlled (independent of PUSCH) via an open loop and a closed loop with Transmission Power Control (TPC) commands transmitted on PDCCH. The power setting formula for PUCCH, as per 3GPP TS , is as follows: For PUCCH, the UE sets the power according to the formula P PUCCH = min {P CMAX, P 0_PUCCH + PL + h(n cqi, n harq ) + Δ FPUCCH (F) + g(i)} [dbm] LZT R1A Ericsson AB

226 LTE L13 Radio Network Functionality Note that for PUCCH, full pathloss compensation is always used, and the resource allocation is always one resource block. The reason for using full pathloss compensation is that different PUCCH users are code multiplexed (CDMA). They may have different path losses, and this could lead to harmful interference between the users if the received powers differ significantly. Further, the function f for the PUCCH always represents accumulation. The closed loop correction δ PUCCH is sent in DL assignments on the PDCCH, or in special TPC_PUCCH messages on the PDCCH. The latter may be configured to be the same as the TPC_PUSCH. pzeronominalpucch P PUCCH = min{p CMAX, P 0_PUCCH + PL + h( n cqi, n harq ) + Δ F_PUCCH (F) + g(i)} P CMAX : max UE power accoring to its class or cell restriction P 0_PUCCH : Target PSD PL : Estimated DL path loss h(n) : = 0, when normal CP is used FPUCCH : PUCCH format offset g(i) : Accumulation function PUCCH : Power control step ( PUCCH TPC command). Input to g(i) Closed loop δ PUCCH included in Downlink Scheduling Control (when present) δ PUCCH for multiple UEs jointly coded and transmitted on PDCCH Used when no Downlink Scheduling Control Figure 5-89: Uplink Power Control PUCCH P CMAX is the configured UE transmitted power defined in 3GPP TS P 0_PUCCH is the target PSD rx, corresponding to P 0_PUCCH from PUSCH. It is set according to the parameter pzeronominalpucch and signaled separately on BCCH System Information Blocks (SIBs). PL is the pathloss estimate and is the same as for PUSCH. h(n) is a PUCCH format dependent value where n cqi corresponds to the number of information bits for the CQI, and n harq the number of HARQ bits. For PUCCH formats 1 (SR), 1a (ACK/NACK) and 1b (ACK/NACK for two CWs), h(n) is always 0. For PUCCH format 2 (CQI), h(n) is 0 for normal cyclic prefix and n cqi <4, but may take on other values otherwise. Δ F_PUCCH (F) is an offset that depends on which information is transmitted on PUCCH. It is provided by higher layers. F corresponds to a PUCCH format. g(i) is the current PUCCH power control adjustment state Ericsson AB 2012 LZT R1A

227 Capacity Management The open loop part controls P 0_PUCCH, and the closed loop part controls g(i). The function works much the same way as for PUSCH, except the TPC commands for PUCCH can be transmitted in the downlink scheduling assignments, and without any uplink grant. δ PUCCH is set according to: TPC command Accumulated δ PUCCH [db] Figure PUCCH Power Setting. So called inter-cell power control is assisted by the X2-based overload indicator. This may be used by the enb as an input to power control and scheduling. P0 for PUSCH and PUCCH statically signaled in SIB2 UL PL estimated from DL PL (same as for PUSCH) h(ncqi, nharq) = 0 except when extended CP and PUCCH format 2 is used ΔF PUCCH is a PUCCH format correction relative to PUCCH format 1a g( i) g( i 1) ( i 4) PUCCH SI SIB2: P0 PUCCH PDCCH: δ P 0 = PSD RXtarget for PUCCH, including correction for non orthogonal CDMA and frequency difference PUCCH P max, PL P PUCCH P PUCCH = min {P CMAX, P0 PUCCH + PL + h(n cqi, n harq ) + Δ F_PUCCH + g(i )} [dbm] Figure 5-91: PUCCH Power Control LZT R1A Ericsson AB

228 LTE L13 Radio Network Functionality ADDITIONAL MAXIMUM POWER REDUCTION Additional Maximum Power Reduction (A-MPR) is a part of LTE Basic. The feature enables operator control of A-MPR for frequency bands 1, 2, 4, 10, 13, 35, and 36. A-MPR is needed to meet additional emission requirements in specific deployment scenarios. UE is allowed to use a lower transmit power Needed to meet additional emission requirements in specific deployment scenarios Applicable for operating bands: 1, 2, 4, 10, 13, 35, and 36 SIB2: additionalspectrumemission NetworkSignallingValue { NS_01, NS_03, NS_04, NS_05, NS_07 } enb Figure 5-92 Additional Maximum Power Reduction A-MPR For the particular operating bands, the enodeb can issue a special Network Signaling (NS) Value to the UE to allow it to apply A-MPR for the uplink output power. This implies that the UE is allowed to transmit with a lower transmit power, i.e. the lower limit (P CMAX_L ) of P CMAX is decreased. The operator can select networksignallingvalue values NS_01, NS_03, NS_04, NS_05, or NS_07. The operator selected networksignallingvalue is mapped on the information element additionalspectrumemission that is transmitted on SystemInformationBlockType 2. Note: NS_01 is default value with meaning "no A-MPR". NS_07 is intended for band Ericsson AB 2012 LZT R1A

229 Capacity Management PUCCH OVERDIMENSIONING PUCCH Overdimensioning is a licensed feature. It is primarily intended for Band 13, where proximity to US public safety bands puts severe restrictions on the handling of uplink transmissions. For band 13 support, the enodeb needs to issue NS_07 signaling to the UE for the UE to apply A-MPR (power back-off), to not interfere with the adjacent Public Safety (PS) band (networksignallingvalues NS_01 is the default meaning no back-off, and NS_07 is the one to use for Band 13). To overcome this power problem, the new PUCCH Overdimensioning feature is needed. This uses the fact that uplink system bandwidth closest to the neighboring bands (that is, at the band edges) is used for PUCCH CQI transmissions. By increasing the size of this region, it is possible to move the "active part" of the PUCCH closer to the middle of the system bandwidth. In this context, the "active PUCCH" part of PUCCH refers to the PUCCH PRBs where CQI and SR resources are actually assigned. Release L11A does not use the spectrum outside of this active part, while releases from L11B can use it for PUSCH. PUCCH active part (CQI and SR resources) moved closer to the middle of the bandwidth Example: Reduce interference, together with A-MPR, with US public safety bands when using operating band 13 pucchoverdimensioning PUCCH Active Part pucchoverdimensioning PUSCH Figure 5-93 PUCCH Overdimensioning The feature is activated by setting the value of pucchoverdimensioning. This value corresponds to the number of physical resource blocks outside of each active PUCCH region. The feature will impact PUSCH performance. One can expect to experience a reduction in peak rate and reduction in cell throughput. Note: The PUCCH overdimensioning feature is optimized for NS_07 for a 10 MHz cell bandwidth on band 13, but the feature can also be used for, for example, band 12 and band 17. LZT R1A Ericsson AB

230 LTE L13 Radio Network Functionality 6 Link Adaptation Uplink link adaptation, i.e., selection of modulation and channel coding, is controlled by the network. The enb measures the uplink channel quality and orders the UE to use a specific modulation and coding scheme (MCS) based on this. Other parameters may also be taken into account, such as UE power headroom, scheduled bandwidth, buffer content and acceptable delay. Rapid interference variations make it difficult to predict the link quality accurately, and select MCS based on such knowledge. Instead, preliminary, MCS selection is based on averaged link quality. Different operating points can then be used. To reach low delay (few retransmissions), a margin to the interference variations can be included. This however leads to limited throughput, as often an unnecessary robust MCS is used. To reach high throughput, a low margin (even negative) is used. This will instead lead to a larger number of retransmissions, and hence a larger delay. The risk of throughput loss or large delays in case of negative margins is reduced by the use of incremental redundancy for retransmissions. Adapts MCS (code rate, QPSK, 16-QAM, 64-QAM) Adaptation based on link quality estimation Used for new transmissions and retransmissions HARQ OPP used as DL quality requirement HARQ OPP is targeted no of tx and resulting BLER BLER is used for channels without HARQ SINR used for UL LA Worst case LA used for initial messages (BCCH, PCH and RA response). This means MCS is chosen to reach cell edge Figure 5-94: Link Adaptation 6.1 Channel prediction In the downlink, Gain to Interference and Noise Ratio (GINR) is used as a measure for channel prediction. It varies due to fading and interference. The slow fading component is tracked and used in link adaptation. GINR can be converted to Signal to Interference and Noise Ratio (SINR) by adding Power Spectral Density (PSD) logarithmically, Figure SINR = GINR + PSD Ericsson AB 2012 LZT R1A

231 Capacity Management Gain to Interference and Noise Ratio (GINR) is independent of tx power and used in order to translate e.g. SINR RS to SINR PDSCH SINR RS = GINR + PSD RS SINR PDSCH = GINR + PSD PDSCH Figure 5-95 What is Gain to Interference Noise Ratio The UE estimates SINR based on the PSD of the downlink Reference Signals (RS) and PSD offset between Physical Downlink Shared Channel (PDSCH) and RS. This SINR is then converted to Channel Quality Indicator (CQI) and reported to the RBS in CFRs. The CQI indicates the radio quality, and is used by the link adaptation function to select the transport format matching the channel conditions. This will lead to improved radio resource utilization. GINR is used for channel prediction SINR = GINR + PSD GINR filtered in order to remove quick fluctuations (fast fading) SINR PDSCH is used for LA CQI -> SINR RS estimated via turbo decoder performance and 3GPP CQI mapping PSD RS,TX GINR = SINR RS PSD RS,TX SINR PDSCH = GINR + PSD PDSCH,TX CQI PSD RS, RX => SINR RS SINR RS => CQI PSD RS,RX Figure 5-96: DL Channel Prediction CFRs contain CQI, Precoding Matrix Indicator (PMI) and Rank Indicator (RI). The latter two are only used if a Multiple Input Multiple Output (MIMO) transmission with spatial multiplexing is present. PMI is only used for closed loop spatial multiplexing, while CQI and RI are used also for open loop spatial multiplexing. L12 only supports tx diversity and open loop spatial multiplexing. CFRs are transmitted either periodically over the Physical Uplink Control Channel (PUCCH) every 40ms or when the RBS triggers one over the Physical Uplink Shared Channel (PUSCH) based on downlink data activity and the age of the earlier received CFR. The RBS performs an adaptive adjustment of the SINR derived from CQI to compensate for errors and mismatches, and fulfills the targeted operating point. LZT R1A Ericsson AB

232 LTE L13 Radio Network Functionality CQI is estimated by the UE. It is up to the UE vendors to decide how to do it according to performance requirements. In the enodeb, CQI is mapped to SINR via the expected turbo decoder performance given the 3GPP definition of CQI values in terms of modulation and code-rate. The RS PSD is removed from SINR to get GINR. Time filtering is applied on GINR to remove fast variations that can not be handled given the CQI feedback rate. GINR + PDSCH PSD give the effective SINR for PDSCH which we use for LA. In the uplink, the channel quality estimate consists of predicted transmitted PSD, PSD TX, uplink Gain (path-loss) estimate, and "Noise+Interference" measurement. The PSD TX is estimated based on UE reported PHR (power headroom report). This is used to estimate uplink gain together with measured PSD RX. "Noise+Interference" is separately estimated per cell. Gain G (path-loss) calculated by Channel Prediction Prediction of PSD TX is based on PHR and BW G = PSD RX PSD TX SINR = PSD TX + G (N+I) SINR is used for LA G is used for PC PSD RX Measure PSD RX Estimate N+I and PSD TX Calculate G SINR = PSD TX + G (N+I) PHR G PSD TX Figure UL Channel Prediction. PH = P CMAX {10logM + P0_PUSCH + α PL + f(i )+ Δ TF } [db] The power headroom (PH) that the UE reports to the enb in a power headroom report (PHR) is defined by: PH = P CMAX {10logM + P 0_PUSCH + PL + f()+ TF } [db] Ericsson AB 2012 LZT R1A

233 Capacity Management The PH is rounded to the closest integer value between -23 to 40dB. P TX P UMAX If UE reaches P UMAX, the PSD TARGET cannot be kept and the power is distributed over the allocated RBs PSD TX N PRB This is indicated by a negative PHR PSD Target This leads to frequency diversity, but also possible problems to maintain SINR TARGET TBS N PRB N PRB Figure 5-98 Power and TBS Relationship Link Adaptation uses link quality measurements to adapt the MCS. The MCSs consist of a modulation constellation (QPSK, 16QAM or 64QAM) and coding rate (ratio of information and coded bits). Link Adaptation is used on new transmissions and HARQ retransmissions. It is not used in the uplink for random access message 3 ("RRC Connection Request"). For channels without retransmission, the quality requirement enforced is Block Error Rate (BLER). For channels with retransmission the quality requirement is HARQ OPP (HARQ Operating Point). HARQ OPP is defined by the targeted number of transmissions and the BLER after the targeted number of transmissions. In the downlink, the selection is channel dependent because Link adaptation may use CQI reports from the UEs to adapt the transmissions to current radio conditions. Before the first CQI message is received, worst case link adaptation is used. This is where it is assumed the UE is at cell edge. Link Adaptation of downlink common channels uses a fixed code rate to ensure these transmissions reach the cell edge. For the uplink, Link Adaptation takes SINR into account. The SINR is based on measurements on the uplink demodulation reference signal. Worst case link adaptation is used until the first power headroom report is received. The following common channels use QPSK modulation: PBCH, PDCCH and PCFICH. PHICH uses BPSK. LZT R1A Ericsson AB

234 LTE L13 Radio Network Functionality 6.2 PDSCH Link Adaptation The quality requirement for this channel is HARQ OPP. The transport format parameters are MCS and TBS (Transport Block Size). The MCS is signaled to the UE in the scheduling assignment. The MCS together with the resource assignment will determine the TBS and coding rate as specified in There are 29 possible MCS values for new transmissions. HARQ OPP MCS signaled to UE in scheduling assignment MCS + resource assignment => TBS and CR MCS Index Modulation Order TBS Index reserved 31 6 Figure PDSCH LA. The maximum TBS depends on the amount of data in the scheduler and capability of the UE. For HARQ retransmissions, three MCS values correspond to the modulation used. The TBS in this case is the same as the initial transmission and not signaled. The supported values of MOD for PDSCH can be QPSK, 16QAM and 64QAM. This corresponds to 2, 4 and 6 bits per modulation symbol respectively. The latter is only possible with a license. Link Adaptation is frequency selective in this release, which means that the selected MCS is based on the estimated channel quality over the assigned resource blocks Ericsson AB 2012 LZT R1A

235 Capacity Management I TBS N PRB Figure TBS table extract. 6.3 PDCCH Link Adaptation Multiple PDCCHs can be transmitted in one subframe, mapped to resource elements on one or several of the first three Orthogonal Frequency Division Multiplexing (OFDM) symbols not being used for reference signals, PCFICH or PHICH. PDCCH REs are divided into a number of Control Channel Elements (CCEs), each containing 36 REs. The REs within a CCE are spread out in frequency and time to achieve diversity. One PDCCH can be mapped to 1, 2, 4 or 8 CCEs, which can belong to different OFDM symbols. Selection of number of CCEs is done based on the same GINR estimate used for PDSCH link adaptation adjusted by an additional fixed margin. LZT R1A Ericsson AB

236 LTE L13 Radio Network Functionality PDCCH is transmitted with QPSK modulation. Convolutional coding 1/3 and rate matching (to fit data onto CCEs) are applied to data on PDCCH. QPSK Convolutional Coding R=1/3 Rate Matcing to fit CCEs 1,2,4 or 8 CCEs, each CCE containing 36 REs Number of CCEs based on GINR (same as PDSCH) A margin (back-off) is added to PDSCH GINR to compensate for different interference scenarios between the two channels Figure 5-101: PDCCH LA 6.4 PUSCH Link Adaptation PUSCH link adaptation needs power headroom reporting to predict the PSD TX of UL transmissions. Power headroom reporting provides the serving RBS with information about the difference between the UE's maximum transmit power and the target power for PUSCH transmission. These reports are sent by the UE if the measured pathloss has changed by a given amount, or the time since the last report is a given amount. This is indicated by the parameters deltapathloss and phrtimer parameters signaled by higher layers. PUSCH LA based on PHR PHR sent if UE pathloss has changed by a fixed amount (deltapathloss) PHR sent if a fixed timer has elapsed since last report (phrtimer) The UE could be prohibited to send the PHR during a fixed amount of time (tphrprohibited) even if pathloss has changed more than a fixed amount pathloss PHR only sent when there are PUSCH allocations Reported value POWER_HEADROOM_0 POWER_HEADROOM_1 POWER_HEADROOM_2 POWER_HEADROOM_3 POWER_HEADROOM_4 POWER_HEADROOM_5 POWER_HEADROOM_57 POWER_HEADROOM_58 POWER_HEADROOM_59 POWER_HEADROOM_60 POWER_HEADROOM_61 POWER_HEADROOM_62 POWER_HEADROOM_63 Measured quantity value (db) -23 PH PH PH PH PH PH PH PH PH PH PH PH 40 PH 40 Note: deltapathloss, tphrprohibeted and phrtimer are not real parameter names Figure 5-102: Power Headroom Report Ericsson AB 2012 LZT R1A

237 deltapathloss deltapathloss deltapathloss Capacity Management Also, there is another parameter tphrprohibited, which is the minimum time between a Power Headroom Report (PHR), even when the pathloss change is fulfilled. PHRs are only sent when there are PUSCH allocations. Note: deltapathloss, phrtimer and tphrprohibited are not real parameter names! The parameters are non-configurable. Path Loss PHR PHR PHR PHR PHR tphrprohibited tphrprohibited tphrprohibited tphrprohibited phrtimer phrtimer phrtimer Figure Triggering of PHR. PUSCH link adaptation differs to PDSCH link adaptation in the following ways: No scheduling grant needs to be transmitted for a retransmission if the allocation is sent in the same frequency and transport format. The PSD is not constant between different allocations. This depends on the number of scheduled blocks, where the upper bounds can be limited by UE power in order to avoid too low SINR. CFR and HARQ Acknowledgements can be multiplexed with uplink data over PUSCH. LZT R1A Ericsson AB

238 LTE L13 Radio Network Functionality MCS Index Modulation Order TBS Index I TBS Redundancy Version SINR & HARQ OPP MCS signaled to UE in scheduling grant MCS + #RBs granted => => TBS and CR I MCS Q m reserved rv idx Figure PUSCH LA. 6.5 PUCCH Link Adaptation Link adaptation is not applicable to PUCCH. 6.6 Power Spectral Density for Power Control of PUCCH and PUSCH The Power Spectral Density (PSD) for power control of PUCCH and PUSCH are controlled by the parameters pzeronominalpusch and pzeronominalpucch. The parameters provide target values for received PSD. Increasing the values of the pzeronominalpusch and pzeronominalpucch parameters result in higher signal power, but will also result in increased intercell interference, which may affect cell edge UE. Decreasing the values of the parameters will have the opposite effect. 6.7 Link Adaptation of Initial Messages. Initial Messages include PBCH, PDSCH carrying BCCH and PCH, and Random Access (RA) response. Here, predictions of GINR and uplink PSD are not known. As a result, MCS selection is made in order to reach cell edge and with the assumption of UE having the lowest P CMAX Ericsson AB 2012 LZT R1A

239 Capacity Management 6.8 PHICH groups The Physical Hybrid-ARQ Indicator Channel (PHICH) is used to transmit HARQ to the UE in response to the reception of UL-SCH transmissions. Eight HARQ transmissions are code multiplexed onto one PHICH group. The number of PHICH groups in a cell is determined by the following equation: Groups = Ceiling [Ng x N DL_RB / 8] where Ng = Maximum Transmission Power The value P RU,Cell, calculated as shown in the following equation, defines the maximum allowed output power from each Radio Unit (RU) that can be used simultaneously by all downlink channels: RU Cell = partofradiopower 100 X Min [HardwareCapability, capacityunitpower, confoutputpower 2 ] Equation 1 Maximum Output Power per RU Used Simultaneously by All DL Channels The maximum transmission power of the cell is available via the read-only parameter maximumtransmissionpower in the EutranCellFDD MO. This parameter shows the maximum available power at the Antenna Reference Point (ARP) for all downlink channels used simultaneously. The reported power level includes the downlink feeder attenuation defined by the dlattenuation parameter. The power level is reported in dbm with a resolution of 0.1dB. 7 MIMO in LTE The LTE specifications support the use of multiple antenna ports at both transmitter (TX) and receiver (RX). MIMO (Multiple Input Multiple Output) uses this antenna configuration. The LTE specifications support up to four antenna ports at the TX side and up to four antenna ports at the RX side (here referred to as 4x4 MIMO configuration). 3GPP release 8 of supports one TX antenna port at the UE, but four RX antenna ports. This leads to that so called Single User MIMO (SU-MIMO) will be supported only in DL (and maximum 4x4 configuration). LZT R1A Ericsson AB

240 LTE L13 Radio Network Functionality Single user MIMO (SU-MIMO) Precoded spatial multiplexing with rank adaption Used in DL Multi user MIMO (MU-MIMO) Tailored for multiple UEs per RB Max one layer per UE Used in UL Transmit diversity Block code based Used in DL Figure LTE transmission types. SU-MIMO increases the data rate for a single user by creating several layers for that user. In UL multi user MIMO can be applied. This means that the basestation uses MU-MIMO to separate different UE transmissions spatially. This leads to that several UEs can be scheduled in the same resource block simultaneously (same frequency, same time). This increases the capacity in the cell. There are seven different transmission modes in LTE, as defined by 3GPP TS Switching between the modes is done by RRC signaling. Mode 1 ( Single antenna port, port 1 ) o One antenna port o Can be used for classical beamforming without precoding feedback Mode 2 ( Transmit Diversity ) o SFBC (Alamouti) o 2 or 4 tx antenna ports Mode 3 ( Open loop spatial multiplexing ) o 2 or 4 tx antenna ports o CQI and RI feedback o Tx schemes: Tx diversity Large delay CDD Mode 4 ( Closed Loop spatial multiplexing ) o 2 or 4 tx antenna ports o CQI, PMI and RI feedback o Tx schemes: Tx diversity CL SM Mode 5 ( Multi User MIMO ) Ericsson AB 2012 LZT R1A

241 Capacity Management o Two UEs can be scheduled in the same RB o Tx schemes: Tx diversity MU-MIMO Mode 6 ( Closed loop spatial multiplexing, single layer ) o As mode 4, but with RI hardcoded to 1 o Tx schemes: Tx diversity CL SM Mode 7 ( Single antenna port, port 5 ) o Can be used for classical beamforming without feedback. The transmission modes defined by 3GPP is shown in Figure The names in brackets are not the formal names. Mode 1 ( Single antenna port, port 1 ) Mode 2 ( Transmit Diversity ) Mode 3 ( Open loop spatial multiplexing ) Tx div & Large Delay CDD Mode 4 ( Closed Loop spatial multiplexing ) Tx div & CL SM Mode 5 ( Multi User MIMO ) Mode 6 ( Closed loop spatial multiplexing, single layer ) Mode 7 ( Single antenna port, port 5 ) Figure GPP LTE DL transmission modes. Every transmission mode can use one or more transmission schemes. Typically, the transmission mode is set-up at session establishment and not changed during the session, while the transmission scheme is dynamically decided every TTI. LZT R1A Ericsson AB

242 LTE L13 Radio Network Functionality The Dual Antenna DL Performance Package feature is controlled by the parameter featurestatedualantdlperfpkg [activated/deactivated]. The license will have a dependency towards MOM configuration of nooftxantennas [0,1,2,4]. Table : Transmission Modes and Transmission Schemes Transmission Mode Transmit Diversity Large Delay Cyclic Delay Diversity TXD (mode 2) OLSM (mode 3) Figure Parameters for MIMO. With TX diversity, DCI (DL Control Information) format 1 is supported. With open loop spatial multiplexing, DCI format 2A is supported. This DCI format includes allocation of two transport blocks with different modulation and coding schemes (MCSes) as well as bitmap based resource allocation. The Dual antenna downlink performance package introduces two different multi antenna transmission schemes to the LTE RAN downlink; Transmit diversity (Transmission mode 2) and Open loop spatial multiplexing (Transmission mode 3). The Transmit diversity mode utilizes an Alamouti (space-frequency block code) scheme in the frequency domain to transmit one codeword on two antennas. The transmit diversity scheme is always rank 1 and does not in itself enhance the data rates. Instead, the gain in using transmit diversity is that it provides a better received signal quality (SINR) at the UE side. The Open loop spatial multiplexing mode supports both rank 1 and rank 2 transmissions. The rank 1 form of this scheme is equivalent to transmit diversity, as described above. The rank 2 form uses large delay cyclic delay diversity (CDD) to transmit two codewords on two antennas by using spatial multiplexing. If the right conditions are fulfilled, this enhances the data rates. Open loop spatial multiplexing in LTE is an adaptive multi-antenna scheme that selects the scheme with the most suitable properties, to optimize the usage of the radio resource. For lower SINR, transmit diversity is the best suited multiantenna scheme. For higher values of SINR, the UE benefits from using spatial multiplexing Ericsson AB 2012 LZT R1A

243 Capacity Management Spatial Mux works better at high SNRs (close to BS) Tx Div improves SNR at cell edges Spectral efficiency [bits/s/hz] Higher peak rate with SM Tx div (rank 1) SM (rank 2) Adaptive tx scheme Better cell egde perf/coverage Close to BS Optimal switching point Cell edge Distance (- SINR) Figure Typical Scenario with MIMO. To adaptively switch the rank of the transmission, as in the Open loop spatial multiplexing mode, channel state information is necessary at the transmitter side. In LTE, the UE estimates the signal quality and decides the most suitable rank for transmission. Information about signal quality and preferred rank is fed back in CQI reports and rank indicators (RI) to the base station. If the Transmit diversity mode is used, rank indicators are not transmitted since the rank is statically set to 1. Precoder Matrix Indicator (PMI) is not necessary for open loop spatial multiplexing since large delay CDD is used. At connection to a cell, the UE is configured to provide channel state information with different contents and periodicity, depending on the transmission mode. If the transmission mode is set to Transmit diversity, the UE is requested to provide channel quality indicator (CQI) reports periodically on the physical uplink control channel (PUCCH). If the transmission mode is set to Open loop spatial multiplexing, the UE is, in addition to CQI, requested to periodically report rank indicators (RI) on PUCCH. The periodicity and contents of the periodic CSI reporting is set via RRC messaging. In addition to this, the UE may also send aperiodic CQI reports together with data on the physical uplink shared channel (PUSCH). These reports are requested by link adaptation algorithm and are configured to contain information on the channel quality. LZT R1A Ericsson AB

244 LTE L13 Radio Network Functionality 7.1 Downlink scheduling and Link adaptation For a UE that is selected for scheduling, the first step is to select a downlink control information (DCI) format. In case the transmission mode is set to Open loop spatial multiplexing, DCI format 2A is selected. If transmission mode is set to Transmit diversity, DCI format 1 is selected. DCI format 2A supports transmission on one or two codewords, whereas DCI format 1 only supports transmission on one codeword. Both formats support resource allocations of type 0, that is, bitmap allocations. Based on the resource allocation types for the DCI format, the DL scheduler then assigns a set of physical resource blocks (PRB) to the UE. This set of PRB resources, together with the preferred number of bits and number of codewords for transmission serves as input to the Transport Format Selection function. The transport format selection keeps track of UE provided rank indicator (RI) feedback. If two codewords are enabled for transmission and the latest indicated RI feedback is two, two transport blocks are allocated. The transport block size for the transmission is provided by the link adaptation function, based on the channel measurements provided by the UE in channel quality indicator (CQI) reports. After reception, the UE feedbacks separate HARQ acknowledgements for each transmitted codeword. The scheduler is therefore implemented to support all combinations of transmission and retransmission on two codewords, such that the usage of the radio resource for new data is optimized. 7.2 Layer 1 processing When the DL scheduler process is finalized, the result is a set of downlink assignments that are provided as input to the Layer 1 processing unit. The process for each downlink assignment differs depending on the number of codewords in use. If the number of codewords is set to one, the Layer 1 processing involves layer mapping and precoding for transmit diversity as described in chapters and of If the number of codewords is set to two, the Layer 1 processing involves layer mapping for spatial multiplexing and precoding for large delay cyclic delay diversity (CDD) as described in chapters and of The output after mapping codewords to layers and precoding is, for both transmission modes, two sequences of complex valued symbols that are mapped to resource elements for transmission on antenna ports 0 and Ericsson AB 2012 LZT R1A

245 Capacity Management DCI Format Selection DL Scheduling DL assignment: - TBS - MCS - PRB Resource DCI Format Resource Allocation PRB Resource # CW Link Adaptation CQI RI # TB L1 Processing Spatial Multiplexing 2 1 Tx Diversity Antenna Port 0 Antenna Port 1 Figure 5-109: Dual Antenna DL Performance Package Flowchart 7.3 Interference Rejection Combining (IRC) IRC maximizes the SINR of the combined signal from two antenna ports. It works in the spatial domain and suppresses interfering signals by weighting down the signal in the directions of the interferers. IRC outperforms MRC (Maximum Ratio Combining) in interference limited systems. When IRC feature is enabled (featurestateirc = true), the RBS will use IRC instead of MRC in interference limited situations. Inter Cell Interference can be suppressed by spatial combining of multiple receiving antennas Increases the Capacity Replaces Maximum Ratio Combining (MRC) Cancellation efficiency varies due to load and number of interferers interferer Figure Interference Rejection Combining (IRC) LZT R1A Ericsson AB

246 LTE L13 Radio Network Functionality IRC is supported for 2, 4 and 8 Rx antenna configuration on PUCCH, PUSCH and from L13A also on PRACH. 8 Parameters 8.1 QoS Parameters Parameter logicalchannelgroupref Description Controls the mapping of QCI to logical channel group dscp Controls the mapping of QCI to DSCP (0 63) priority QCI related priority. Default values in accordance to 3GPP TS Scheduling Parameters Parameter configurablefrequencystart dlinterferencemanagementactive ulinterferencemanagementactive schedulingstrategy abspriooverride nrofsymbolspdcch Description Specifies the start frequency offset as a percent of the bandwidth available. Controls whether Random Frequency Allocation is enabled or disabled. Controls whether Random Frequency Allocation is enabled or disabled. Controls whether Round Robin or Proportional Fair is used as the scheduling algorithm for an RBS. Indicates if the DRBs of the chosen LCG are subject to absolute prio override. This applies to both downlink and uplink. Read only parameter. Specifies the number of OFDM symbols used for PDCCH. ulinterferencemanagementactive ulfrequencyallocationproportion noofpucchsrusers noofpucchcqiusers The values 1 3 indicate that a fixed number of OFDM symbols for control are used. The value 0x7FFF indicates that the number of OFDM symbols for control is fixed but based on the system downlink bandwidth. If a cell has a bandwidth 5 MHz then 2 OFDM symbols are used. Otherwise, 1 OFDM symbol is used. Controls whether Random Frequency Allocation (ICIC) is enabled or disabled. Frequency resources that is allocated in UL expressed as a percentage of the configured bandwidth Determines the number of scheduling request resources available on PUCCH. Determines the number of CQI resources available on PUCCH. 8.3 Minimum rate proportional fair scheduler parameters Parameter featurestatepfs keyidpfs servicestatepfs Description Indicates whether the licensed feature UL and DL Proportional Fair Scheduling with minimum rate scheduling is ACTIVATED or DEACTIVATED. The license key ID for the UL and DL Proportional Fair Scheduling with minimum rate scheduling feature. Indicates if the UL and DL Proportional Fair Scheduling with minimum rate scheduling feature is operable or inoperable Ericsson AB 2012 LZT R1A

247 Capacity Management PfsId licensestatepfs ulminbitrate dlminbitrate schedulingalgorithm relativepriority The value component of the Relative Distinguished Name (RDN). Indicates whether the licensed state of feature UL and DL Proportional Fair Scheduling with minimum rate scheduling is ENABLED or DISABLED, if a valid license key is installed for the feature. The scheduler will attempt to achieve minimum bit rate for all bearers before giving any user a higher rate. This parameter will only be used if one the five Proportional Fair scheduling algorithms has been selected. The scheduler will attempt to achieve minimum bit rate for all bearers before giving any user a higher rate. This parameter will only be used if one the five Proportional Fair scheduling algorithms has been selected. Specifies which scheduling algorithm is to be used for a certain QCI. Relative priority for Proportional Fair Scheduling 8.4 UL frequency selective schduling Parameter featurestateulfss keyidulfss servicestateulfss UlFssId licensestateulfss resourceallocationstrategy srsallocationstrategy ulsrsenable Description Indicates whether the licensed feature Uplink Frequency-Selective Scheduling is ACTIVATED or DEACTIVATED. The value of the attribute is irrelevant when no valid license key is installed for the feature. The license key ID for the Uplink Frequency-Selective Scheduling feature. Indicates if the Uplink Frequency Selective-Scheduling feature is operable or inoperable. The value component of the Relative Distinguished Name (RDN). Indicates whether the licensed state of the feature Uplink Frequency-Selective Scheduling is ENABLED or DISABLED, whether a valid license key is installed for the feature. Defines the resource allocation strategy of the QoS Class Identifier (QCI). Changes to this parameter take effect only if the licenses for Uplink Frequency-Selective Scheduling and QoS-Aware Scheduler are present and the features are active. If set to ACTIVATED, an attempt is made to allocate sounding for a UE. If several Data Radio Bearers are setup towards the UE with different QoS configurations, and the QoS configurations have different QCI parameters, an algorithm using the priority parameter in the QoS configuration will resolve which QoS configuration that will define sounding. This parameter controls whenever sounding shall be enabled or not. When enabled, cell resources are configured for sounding. 8.5 Power Control Parameters Parameter Description pzeronominalpusch pzeronominalpucch maximumtransmissionpower partofradiopower confoutputpower pmaxservingcell Specifies the nominal component of the UE transmit power for PUSCH. Specifies the nominal component of the UE transmit power for PUCCH. Maximum possible power at the antenna reference point for all downlink channels added together used simultaneously in a cell. The requested part of the total radio power per antenna connector that should be allocated for the cell. The requested maximum sector power. The value represents the sum of the power for all antenna connectors used by the sector. Maximum power to be used in the cell. If absent the UE applies the maximum power according to the UE capability. LZT R1A Ericsson AB

248 LTE L13 Radio Network Functionality capacityunitoutputpower Defines the unit for licensed capacity. Range: Preamble Parameters Parameter rachrootsequence Description First root sequence number for RACH(1) preamble generation RACH root sequence is broadcast as a part of system information distribution and used for preamble detection. The definition for logical root sequence number can be found in 3GPP TS cellrange Defines the maximum distance from the RBS where a connection to a UE can be setup or maintained, or both 8.7 MIMO Parameters Parameter featurestatedualantdlperfpkg nooftxantennas Description A parameter of the MO DualAntDlPerfPkg. Activates or deactivates the licensed feature Dual-Antenna Downlink Performance Package. The value of the attribute is irrelevant when no valid license key is installed for the Dual-Antenna Downlink Performance Package. The number of antennas that can be used for downlink transmission. A parameter of the MO EUtranCellFDD. Valid values: 0,1,2 and 4, where 0 is the maximum number of antennas. 8.8 IRC Parameters Parameter featurestateirc licencestateirc servicestateirc Description Activates or deactivates the licenced feature IRC Indicates the status of the licenced feature InRC, ENABLED or DISABLED. Indicates whether the feature IRC is operable, and whether the feature is providing service. 8.9 End user bit rate shaping Parameters Parameter rateshapingactive featurestaterateshaping keyidrateshaping licensestaterateshaping RateShapingId Description Indicates whether the End-User Bitrate Shaping feature functionality is activated in this cell. The default value is false. Activates or deactivates the End-User Bitrate Shaping licensed feature. The attribute value is irrelevant when a valid license key is not installed for the feature. The default value is DEACTIVATED. Product identity of the key for the End-User Bitrate Shaping licensed feature. It ranges between [0,30]. Indicates the licensed status of the End-User Bitrate Shaping feature, ENABLED or DISABLED. The value is ENABLED when a license key is installed. The value component of the RDN Ericsson AB 2012 LZT R1A

249 Capacity Management servicestaterateshaping Indicates whether the feature End-User Bitrate Shaping feature is operable or inoperable 8.10 Delay Based Scheduling Parameters Introduced Parameters in MO DelayBasedSchedulingAndSabe Parameter featurestatedbsandsabe keyiddbsandsabe licensestatedbsandsabe servicestatedbsandsabe DBSAndSabeId Parameter ServiceType pdb Parameter ServiceType pdb Description Indicates whether the licensed feature Delay based scheduling and SABE is ACTIVATED or DEACTIVATED. The value of the attribute is irrelevant when no valid license key is installed for the the Maximum Cell Range feature. The license key ID for the delay-based scheduling and SABE feature. Indicates whether the licensed state of feature delay-based scheduling and SABE is ENABLED or DISABLED, whether a valid license key is installed for the feature. Indicates if the feature delay-based scheduling and SABE is operable or inoperable. The value component of the Relative Distinguished Name (RDN). Introduced Parameters in MO QciProfileOperatorDefined Description Indicates the service that the bearer is used for. The contribution from enodeb to the Packet Delay Budget (PDB) for a QCI. Description Introduced Parameters in MO QciProfilePredefined Indicates the service that the bearer is used for. The contribution from enodeb to the Packet Delay Budget (PDB) for a QCI Parameters for Relative Priority Scheduling Introduced Parameters to QciProfileOperatorDefined MO by Relative Priority Scheduling Parameter relativepriority Description Defines the bit rate proportion of the QCI in relation to other QCIs. This parameter can be configured with a valid range of Default value is 1. Note that this parameter is ignored if the feature is not enabled DRX Configuration Parameters Parameter Value Description ondurationtimer 2 ms Specifies the number of consecutive PDCCH subframes at the beginning of a DRX Cycle drxinactivitytimer 100 ms Specifies the number of consecutive PDCCH subframes after successfully decoding a PDCCH indicating an initial UL or DL user data transmission for this UE, in 100 ms drxretransmissiontimer 2 ms Specifies the maximum number of consecutive PDCCH subframes for as soon as a DL retransmission is expected by the UE longdrxcycle 40 ms Specifies the periodic repetition of active time resulting from a started On LZT R1A Ericsson AB

250 LTE L13 Radio Network Functionality Duration Timer followed by a possible period of inactivity The parameter is applied when the UE follows the long DRX cycle. shortdrxcycle 20 ms Specifies the periodic repetition of active time resulting from a started On Duration Timer followed by a possible period of inactivity The parameter is applied when the UE follows the long DRX cycle. drxshortcycletimer 20 ms Specifies the number of consecutive subframes the UE must follow the short DRX cycle after the DRX Inactivity Timer has expired Parameter drxactive Description Determines whether the DRX functionality of the cell is activated MO: EUtranCellFDD featurestatedrx Indicates whether the licensed feature DRX is ACTIVATED or DEACTIVATED MO: Drx keyiddrx The license key ID for the Efficient DRX/DTX for Connected UE feature MO: Drx licensestatedrx Indicates whether the licensed state of the Efficient DRX/DTX for Connected UE feature is ENABLED or DISABLED, and whether a valid license key is installed for the feature MO: Drx servicestatedrx Indicates if the feature DRX is operable or inoperable MO: Drx DrxId Value component of the RDN (1) MO: Drx 8.13 Parameters TTI Bundling Parameter featurestatettibundling keyidttibundling licensestatettibundling servicestatettibundling Description Activates or deactivates the licensed feature TTI Bundling. The value of the attribute is irrelevant when no valid license key is installed for the feature. The license key ID for the TTI Bundling feature. The license status of feature TTI Bundling, ENABLED or DISABLED. The value is ENABLED when a license key is installed. Indicates if the feature TTI Bundling is operable or inoperable. TtiBundlingId The value component of the RDN (1) Ericsson AB 2012 LZT R1A

251 Capacity Management 6 Capacity Management Objectives After this chapter the participants will be able to: Describe the purpose and function of the Capacity Management 1. Describe the interaction between the Monitored System Resources (MSRs) and the different algorithms 2. Explain the static and dynamic MSRs 3. Explain Admission Control 4. Explain Congestion Control 5. Explain the interaction with QoS Figure 6-1: Objectives of Chapter 6 LZT R1A Ericsson AB

252 LTE L13 Radio Network Functionality 1 Capacity Management Introduction Capacity Management aims to control the load in the LTE RAN. The purpose of bearer admission control is to decide, at the time of EPS bearer setup or modify, whether or not a new UE or E-RAB should be admitted to the network. A new UE or E-RAB should be admitted if and only if its quality of service can be satisfied without jeopardizing the QoS constraints of existing E-RABs there are sufficient system resources so that the new UE or E- RAB does not cause overload A new request should only be admitted if there are sufficient system resources so that the new UE or E-RAB does not cause overload. Admission control may also be used to enforce license limitations on the maximum number of active users. For an illustrated overview of Capacity Management functions see following figure. Admission Control at: RRC Connection Set-up EPS Bearer Set-up Handover Request EPS Bearer Modify Admission Control Monitored System Resources Handling Figure 6-2: Capacity Management Ericsson AB 2012 LZT R1A

253 Capacity Management Monitored system resources (MSR) The admission and congestion control algorithms are defined based on a set of metrics referred to as monitored system resources. A monitored system resource (MSR) is a metric that represents a limited system resource that needs to be explicitly monitored to ensure correct and efficient operation of the admission and congestion control procedures. An MSR may be based on abstractions in order to simplify its definition and facilitate the processing. An MSR is either static or dynamic. A static MSR is completely controlled by the admission control algorithms and can therefore be correctly tracked using dead reckoning. The MSR value only changes at the time of admission (if the admission request is granted) and release. A dynamic MSR, on the other hand, is not directly controlled by the admission control algorithm. For a dynamic MSR, measurements are required to keep track of the current MSR value. Moreover, a dynamic MSR may also require congestion control to keep the value at an acceptable level. The difference between the two MSR types is illustrated in figure below. MSR value Static MSR time MSR value Dynamic MSR time Figure 6-3 Static and Dynamic Monitored System Resources The admission and congestion control procedures must directly or indirectly manage the resource utilization related to UE and bearer context storage so that overload is avoided with respect to the limitations in memory and processing capacity. 2 Admission control The purpose of admission control is to control the admission of UEs and E-RABs in such a way that requested QoS can be achieved for the UE or E-RAB seeking admission, as well as for the UEs and E-RABs previously admitted. LZT R1A Ericsson AB

254 LTE L13 Radio Network Functionality Admission control also reserves a share of system resources to be used only for privileged access, for example emergency calls. By reserving system resources, the accessibility and mobility success rate increases for UEs and bearers assigned with privileged access. To accomplish the requested Quality of Service for UEs and E-RABs, Admission Control rejects UE and E-RAB requests on the S1/X2 interfaces based on whether the RBS system resources are congested. Also, different UEs (or bearers) can be given different priorities (ARP Allocation Retention Priority). This enables the operator to prioritize and protect certain services in case of resource shortage. Lower priority UEs (or bearers) can be released (pre-empted) in order to serve UEs that use more important services. The following types of admission control are used in the LTE basic system: Emergency call Prioritization. A share of system resources to be used only for privileged access, for example emergency calls Transport Network Admission Control. This introduces admission thresholds for a number of critical system resources in order to secure that admitted Guaranteed Bit Rate (GBR) bearers receives acceptable QoS, and to protect the GBR bearers already defined in the system. The following monitored system resources and related parameters are used in LTE, for transport and radio network, respectively: Transport network: dltransnwbandwidth, ultransnwbandwidth Transport network bandwidth, DL and UL respectively dlnongbrratio, ulnongbrratio The wanted resource utilization ratio of non-gbr bearers on transport network and cell levels in congested situations, DL and UL respectively resourcetype Indicates if the resource type of the QCI is GBR or non-gbr paarpoverride ARP level received from EPC. enb interprets ARP as the identifier for privileged access Radio network: Admission Control considers the extra cost for having GBR bearers with absolute priority (abspriooverride) Figure 6-4 Monitored System Resources related parameters The transport network admission control ensures that the sum of GBR (from each E-RAB request in S1AP/X2AP procedures) of the admitted E-RABs does not exceed the configured admission thresholds Ericsson AB 2012 LZT R1A

255 Monitored System Resource Capacity Management A threshold level - the OverLoad Control (OLC) level - is defined for each MSR instance. The admission control algorithm uses this threshold level as the upper limit for accepted MSR values at admission. Admission requests must be denied if the MSR value would go above the specified OLC level. Admission Control is regarded as a basic feature that must be enabled in order to protect the system from overload. This feature cannot be switched GBR2 OLC GBR1 off. Applies for GBR users with absolute priority (abspriooverride) Figure 6-5 Radio Admission Control Time LZT R1A Ericsson AB

256 LTE L13 Radio Network Functionality 2.1 Privileged access Admission control reserves a share from the RBS system resources to be used for privileged access in order to increase accessibility and mobility success rate. A UE that is considered for privileged access can be considered for use of reserved resources. If any resource is congested, the UE request is rejected. The following resource requests are considered to have privileged access and they are allowed to use reserved resources: RRC Connection Request with establishmentcause = emergency RRC Connection Request with establishmentcause = highpriorityaccess Handover request where any bearer have an ARP level equal to paarpoverride A bearer request (initial ctxt setup/erab setup/incoming handover) with ARP level equal to paarpoverride Figure 6-6 Privileged access Admission Requested NO Priviliged access? YES YES Non-reserved resources available? NO YES Reserved or Non-reserved resources available? NO Admission Granted Admission Denied Admission Granted Admission Denied Figure 6-7 TN Privileged access For a bearer that is considered for privileged access, reserved resources can be considered for resource allocation. If any resource is congested, the bearer request is rejected. Consideration of transport network resources is only applicable for bearer requests. Admission control ensures that the sum of GBR (from each E-RAB request in S1AP/X2AP procedures) of the admitted E-RABs does not exceed the configured admission thresholds Ericsson AB 2012 LZT R1A

257 Monitored System Resource Monitored System Resource Capacity Management 100% dltransnwbandwidth ultransnwbandwidth GBR2 Reserved for GBR users with privileged access OLC dltransnwbandwidth *(1-dlNonGbrRatio) ultransnwbandwidth *(1-ulNonGbrRatio) GBR1 Time Figure 6-8: TN Admission Control (non-privileged access) 100% GBR2 dltransnwbandwidth ultransnwbandwidth Reserved for GBR users with privileged access GBR1 OLC dltransnwbandwidth *(1-dlNonGbrRatio) ultransnwbandwidth *(1-ulNonGbrRatio) Time Figure 6-9: TN Admission Control (privileged access) The following features are related to Admission Control: Dynamic GBR admission control Differentiated admission control LZT R1A Ericsson AB

258 LTE L13 Radio Network Functionality 3 Dynamic GBR Admission Control Dynamic GBR Admission Control optional feature ensures that the number of bearers that can be supported while maintaining desired Quality of Service is not exceeded for GBR bearers that are typically used for voice and video services with real-time delay requirements. The GBR capacity in terms of number of bearers is autonomously adjusted to differences in cell sizes or other differences in radio conditions. Optional Feature Protects the existing GBR bearers - By limiting the admission of new GBR bearers Can be used to reserve resources for non-gbr traffic Transmission resources resembles dynamic resources but are static Figure 6-10: Dynamic GBR Admission Control When the capacity is limited by air interface resources or licensed baseband capacity, this feature gives a "service blocking over service dropping" behavior which is typically desired for services using GBR bearers. 3.1 Description At high load, bearers that are already admitted into the system are protected from quality degradation and ultimately dropping, by blocking new bearers from entering the system which would increase the load even further. This gives improved retainability at high load for services using GBR bearers. As the feature tries to limit the amount of air interface and baseband resources used by GBR bearers, it implicitly reserves a minimum amount of resources for non-gbr bearers. This gives protection also to the quality of non-gbr bearers at a level decided by operator configuration. At a minimum, the feature prevents non-gbr traffic from starvation that could otherwise happen since GBR traffic is typically prioritized over non-gbr traffic in the scheduler. The feature monitors all resources handled by the air-interface scheduler which could possibly be a capacity bottleneck and starts to limit the load at an operator configurable level. This ensures smooth and fast introduction and easy operation with minimal effort/cost Ericsson AB 2012 LZT R1A

259 Capacity Management In the following figure the flow diagram explains how this feature handles bearer requests. Monitored System Resources (MSR) Figure 6-11: Flow diagram of feature Load Max capacity Admission Threshold Block new accesses Allow incoming mobility Allow privileged access Non-GBR capacity Allow all requests GBR capacity Figure 6-12: Model of the dynamic Monitored system resource (MSR) Dynamic GBR admission control is performed by monitoring the use of certain system resources, referred to as monitored system resources (MSR), and allowing the setting up of new guaranteed bit rate (GBR) bearers only if the MSR usage by GBR bearers does not exceed a configured threshold. These monitored system resources are scheduler-handled resources and include hardware and software resources as well as license limited resources on both cell and RBS level. LZT R1A Ericsson AB

260 LTE L13 Radio Network Functionality The MSRs include resources such as the usage of physical resource blocks (PRB), scheduling elements (SE), control channel elements (CCE) and baseband capacity. The use of the MSRs is measured during a measurement period after which a filtered value for the MSR usage is calculated. This filtered MSR usage value is updated after each new measurement period. The MSRs considered in the decision of accepting or rejecting admission control are: DL GBR PRB utilization per cell GBR PRB utilization per RBS GBR SE utilization per cell GBR SE utilization per RBS GBR L2 bits utilization per RBS GBR CCE utilization UL GBR PRB utilization per cell GBR PRB utilization per RBS GBR SE utilization per cell GBR SE utilization per RBS GBR L2 bits utilization per RBS GBR CCE utilization Figure 6-13: MSRs for dynamic GBR admission Admission can be rejected by anyone of the MSRs. Each one of them admits or rejects admission of a bearer independently of each other. The GBR threshold is common for all MSRs and controlled, for the downlink and uplink, by the parameters dlnongbrratio and ulnongbrratio introduced in Admission Control. The threshold is given as a ratio and can be expressed as 1 - dlnongbrratio for the downlink and 1 - ulnongbrratio for the uplink Ericsson AB 2012 LZT R1A

261 Capacity Management Resource usage is monitored and reported during measurement periods At the end of measurement period a filtered value of resource usage is computed Resource usage is reported for both initial transmissions and retransmissions for GBR bearers resource usage reports MSR Load measurement period Figure 6-14: Dynamic GBR load Measurement Time The GBR load on the dynamic MSRs is not considered for handover and privileged access. Privileged access is described in more detail in Admission Control. 4 Differentiated Admission Control Differentiated admission control introduces the possibility to support Allocation Retention Priority (ARP) as defined by 3GPP in order to provide prioritized access to the system for certain subscribers. This feature uses pre-emption of lower priority UEs to allow higher priority UEs to connect to the system when available resources are scarce, i.e. when the current resource utilization is above the admission limit. Without this feature all new accesses or bearers would be rejected when the admission limit is exceed and no prioritization is possible. The UE s priority and the treatment of a specific bearer are controlled by the Allocation Retention Priority (ARP) parameter that is sent by the MME to RAN. Relevant features are needed in EPC in order to set the ARP according to operator policies. The ARP contains three information elements; Priority (1=highest priority and 15=lowest priority), Pre-emption Vulnerability Indicator (PVI) (values true or false) and Pre-emption Capability Indicator (PCI) (values true or false). A bearer that is "capable" can pre-empt any bearer that is "vulnerable" provided that it has higher priority. LZT R1A Ericsson AB

262 LTE L13 Radio Network Functionality In case of resource limitations, the feature may free up resources from lower priority UEs or bearers to allow access for higher priority UEs or bearers The feature is optional and under license control ARP - Priority (1-15, 15 lowest priority) - Pre-emption Capability Indicator (PCI) (true/false) - Pre-emption Vulnerability Indicator (PVI) (true/false) Figure 6-15: ARP based Admission Control feature NO Request eg. Initial Access, E-RAB Setup, Incoming handover Resource shortage? YES License, HW, SW limits, or admission control functions Are there UEs to pre-empt? YES NO Admit Pre-empt UEs Admit Reject Figure 6-16: Admission Control Resource Request In L12A pre-emption is done on RRC connection / UE level which means that prioritization between different subscribers is possible while prioritization between different services used by the same subscriber is not possible as it would require pre-emption on bearer level. Bearer pre-emption capabilities are added in L12B. The prioritization capability that is provided by this feature relies entirely on the possibility to pre-empt lower priority users to make room for higher priority users. In other words, the accessibility of higher priority users is kept high at the expense of degraded retainability for lower priority users. In 3GPP UE pre-emption is not addressed as it is anticipated that at least the default bearer have PVI set to "not pre-emptable", and thus it would not be possible to pre-empt a complete UE context. As it is allowed to release bearers and UEs that are inactive this feature treats bearers that have been inactive for certain time as if their PVI is set to "pre-emptable" Ericsson AB 2012 LZT R1A

263 Capacity Management The following principles apply in order to find UEs to pre-empt, see Figure A candidate UE can only pre-empt a served UE: if any of the candidate UE s E-RAB(s) has the PCI= may trigger preemption that has all its E-RAB PVI = pre-emptable and/or that all non-gbr E- RABs have been inactive longer than a pre-defined time If the candidate UE s highest priority bearer has higher priority than the served UE s highest priority bearer * The E-RAB resources required by the candidate UE may pre-empt one or several served UEs Figure 6-17: UE pre-emption algorithm LZT R1A Ericsson AB

264 LTE L13 Radio Network Functionality In the figure below we can see a candidate UE (UE1) requesting admission to a cell where there is high load. You can see how the network prioritizes the new user over the already served user (UE2) in terms of ARP. Admission reject? UE2 Bearer ARP PCI= may trigger PVI= preemptable #1 5 Yes Yes UE1 Bearer ARP PCI= may trigger PVI= preemptable #2 8 No Yes #1 #2 3 9 Yes No Yes Yes Resource shortage! Served user: - Active or Inactive UE1 Candidate: - HO or Non-HO Figure 6-18: Admission Control with Pre-emption UE2 Pre-empted! Admission Control will check if any of the MSRs are overloaded and consider if an admission reject should be triggered. If there is a UE with lower priority (higher ARP) that also is pre-emption vulnerable, then this UE (UE2) will be preempted (released) and the requesting UE (UE1) can be admitted. 5 Differentiated Admission Control with UE and bearer pre-emption This feature was first introduced in L12A and it has been enhanced in L12B. In L12A pre-emption is done only on RRC connection / UE level which means that whenever any bearer level resource is congested it may result in UE preemption. The feature is enhanced in L12B with bearer level pre-emption so it is not needed to pre-empt a UE when resource shortage on bearer level appears. The treatment of a specific bearer is controlled by the Allocation Retention Priority (ARP) parameter that is sent by the MME to RAN. Relevant features are needed in EPC in order to set the ARP according to operator policies Ericsson AB 2012 LZT R1A

265 Capacity Management The optional feature Differentiated Admission Control covers basic mechanisms required to admit or reject requests for UE and bearer level resources in RBS based on ARP Priority Level, Pre-emption Vulnerability, and Pre-emption Capability received from the core network. In addition to UEs, also individual E-RABs can be pre-empted in this release in case of resource shortage. Before any pre-emption attempts based on ARP settings, the enodeb in L12B also considers release of inactive UEs at the time of resource requests (Early UE Pre-emption). A UE is then eligible for Early UE Pre-emption when it has been inactive for longer than the minimum of the configurable range of the MOM parameter tinactivitytimer Feature overview The prioritization capability that is provided by this feature relies entirely on the possibility to pre-empt lower priority bearers to make room for higher priority bearers. In other words, the accessibility of higher priority bearers is kept high at the expense of degraded retainability for lower priority bearers. Note also that the use of pre-emption may be subject to regulatory requirements in certain markets. When RBS resources are congested admission control rejects further nonprioritized E-RAB requests. Admission control pre-empts less prioritized services to give up resources to more prioritized services. Priority of E-RAB given by Allocation and Retention Priority (ARP). LTE L12B release improves pre-emption granularity Pre-emption of bearers (previously only UE pre-emption). The system releases unused but allocated static resources. UE level resource shortage (number of licenses) UE Pre-emption E-RAB level resource shortage (e.g. any GBR related resource) Bearer pre-emption The feature exploits a procedure Early UE Pre-emption to identify unused but still allocated resources Figure 6-19: Differentiated Admission Control Resource shortage may appear at initial access, handover or E-RAB Setup/Modification such that the resource request is rejected by Admission Control. The pre-emption of resources is then initiated, followed by yet another (but no more) request for the previously rejected resource request. LZT R1A Ericsson AB

266 LTE L13 Radio Network Functionality The feature exploits a procedure Early UE Pre-emption to identify unused but still allocated resources explained in following pages. Pre-emption is based on the following components of ARP as defined in TS : Priority Level (valid range 1..15) Pre-emption Capability (shall not trigger pre-emption, may trigger pre-emption) Pre-emption Vulnerability (not pre-emptable, pre-emptable) TS assigns special meaning to ARP Priority level 15: it shall be interpreted as Pre-emption capability set to shall not trigger pre-emption and Pre-emption vulnerability not preemptable Release of unused resources based on inactivity (Early UE Pre-emption) involves no consideration of ARP Configuration A UE is eligible for Early UE Pre-emption when it has been inactive for longer than the minimum of the configurable range of the MOM parameter tinactivitytimer. All UEs inactive longer than tinactivitytimer are released by enb (as part of previous releases software releases), the opearator already has the tinactivitytimer configurable in MOM. Still, inactive UEs represent allocated but unused resources. But there is now a new aspect: before any ARP-based pre-emption Early UE pre-emption can be applied to free up resources UEs are eligible for early UE Pre-emption when inactive for longer than the minimum of the configurable range of tinactivitytimer (mintinactivitytimer) Note: Setting tinactivitytimer to mintinactivitytimer means no UEs eligible for Early UE pre-emption Ericsson AB 2012 LZT R1A

267 Capacity Management Early UE Preemption triggered Any UE inactive for longer than mintinactivitytimer? Yes Release No Ues eligible for Early UE Pre-emption All Ues released 0 mintinactivitytimer tinactivitytimer Inactivity Time Figure 6-20: Early UE Pre-emption When admission control denies an E-RAB due to lack of resources, pre-emption of existent E-RAB is then is initiated. After pre-emption is concluded, E-RAB can yet another time (but no more) check if resources are available. ARP-based Bearer Pre-emption follows the criteria listed below: Only E-RABs with ARP set to Pre-emptable pre-empted ARP Priority level of pre-empted E-RABs lower than requesting E-RAB GBR E-RABs are only allowed to pre-empt GBR E-RABs If the E-RAB selected for pre-emption is last E-RAB of the UE, the UE as a whole is released (enb makes no assumption about default bearer; if a default bearer is gone, the core network will notice it and respond accordingly) LZT R1A Ericsson AB

268 LTE L13 Radio Network Functionality E-RAB resource request Does ARP of E-RAB indicate May trigger pre-emption? Yes No Try Early UE Preemption Any Early UE preemption? Yes No preemption No Try to Pre-empt E-RABs to accommodate request based on ARP Pre-emption concluded No Any pre-emption? Yes Figure 6-21: Bearer Pre-emption UE Pre-emption is relevant for UE-level resource, lets say due to shortage (number of licenses) requested when an RRC Connection is established. Since no ARP is visible at the point of RRC Connection Establishment (not visible until an E-RAB has been established for the UE), how to evaluate a UE compared to already admitted UEs? The UE can temporarily use a license resource so it can establish an E-RAB with a known ARP. After the UE has been evaluated in terms of ARP it either departs from the system (not high enough ARP) and returns its temporarily allocated resource or stays in the system and keeps its temporarily allocated resource ( high enough ARP), since it was able to pre-empt other resources. ARP-based UE Pre-emption assumes: Only UEs with all E-RABs having ARP set to Pre-emptable can be pre-empted Ericsson AB 2012 LZT R1A

269 Capacity Management To be pre-empted, the most prioritized E-RAB of the UE shall be less important than the most prioritized E-RAB of the requesting UE The UE selected for pre-emption shall be such that its most prioritized E-RAB has the lowest possible priority level Connection request Any E-RAB with May trigger pre-emption? Yes No Try Early UE Preemption Any Early UE preemption? Yes No No preemption Try Pre-empt UE based on ARP Pre-emption concluded No Any pre-emption? Yes Figure 6-22: UE Pre-emption EPC needs to be aware of the strict use by enb of the ARP settings (as explained in TS ). 6 Dynamic QoS modification This feature was first introduced in L12B. The treatment of each EPS bearer is controlled by QoS parameters that are sent by the MME to RAN. Relevant features are needed in EPC in order to set QoS parameters and to trigger dynamic changes according to operator policies Feature Description Dynamic OoS modification makes it possible to dynamically change any QoS parameter for one or more EPS bearers. LZT R1A Ericsson AB

270 LTE L13 Radio Network Functionality Dynamic QoS Modification is a feature that allows the operator via EPC to modify all parameters within the QoS Profile for an existing bearer What it is: Dynamic QoS Modification is a feature that allows the operator via EPC to modify all parameters within the QoS Profile for an existing bearer Benefits: The feature makes it possible for EPC to change the behaviour for a bearer and so, change the end user experience It also enables for an operator to introduce a fair usage policy by reduce priority and peak rate for a subscriber that has exceeded his monthly quota Figure 6-23: Dynamic QoS Modification The feature makes it possible for EPC to change the behavior for a bearer and so, change the end user experience It also enables for an operator to introduce a fair usage policy by reduce priority and peak rate for a subscriber that has exceeded his monthly quota. Different operators have different strategies and needs but below are some examples for use cases : Enforcement of fair usage policies. A Mobile Broadband subscriber exceeds his monthly quota and it is desirable to reduce the priority (QCI) and/or peak bit rate (AMBR) of that subscriber Dynamic change of a users priority based on jurisdiction (position) or on request by an operator e.g if an emergency situation is declared Addition of new services (Service Data Flows) to existing bearers. This could change the GBR of an GBR bearer or the QCI of a non-gbr bearer as well as the ARP of the bearer Ericsson AB 2012 LZT R1A

271 Capacity Management The Dynamic QoS Modification on S1AP makes possible for the EPC to modify the QoS Profile of already established E-RABs for a given UE The feature involves the admission control functionality in enb and in the case of resource shortage also the bearer pre-emption functionality Depending on what parameters that needs to be modified the feature might involve the UE in the modification procedure Figure 6-24: Feature Overview The new optional feature Dynamic QoS modification introduces support for the S1AP E-RAB Modify procedure, making it possible for the EPC to modify the QoS Profile of already established E-RABs for a given UE Configuration The QoS Profile, presented below, consists for every parameter of an uplink (UL) and downlink (DL) component. UE AMBR is valid for the UE whilst all other parameters are valid per E-RAB. GBR non- GBR QoS Class Identifier (QCI) Determines which bearers are categorized as GBR and witch are categorized as non-gbr. Also used as a reference to access node-specific parameters for the bearer Allocation and Retention Priority (ARP) Decides whether a bearer setup/modification request can be accepted or needs to be rejected in case of resource limitation. Also used to to decide which bearer(s) to drop during exceptional resource limitation Guaranteed Bit Rate (GBR) Defines the minimum bit rate which can be expected to be available to the bearer when required Maximum Bit Rate (MBR) Defines the maximum bit rate which can be expected to be available to the bearer when required UEAggregate Maximum Bit rate (UE AMBR) Defines the maximum allowed throughput for a UE based upon the sum of all its non-gbr bearer. Despite from the other QoS Profile parameters UE AMBR is on UE level and not on RAB level Figure 6-25: QoS Profile The feature is dependent on implementation in both EPC and in the UE. Ericsson MME has already implemented the E-RAB Modify procedure. The E-RAB Modify procedure in enb is able to work with any non-ericsson MME that fulfills the 3GPP standard regarding E-RAB Modify, i.e. There are no Ericsson internal solutions. LZT R1A Ericsson AB

272 LTE L13 Radio Network Functionality End-user bit rate shaping in downlink is a feature that is not implemented in enb L12B which means that modifications of UE AMBR DL and MBR DL will have no effect in enb. The optional feature Dynamic QoS Modification requires that license Dynamic QoS modification is enabled and that the feature is activated. If not, all E-RAB modify attempts will be rejected. 7 Operator defined QCI This feature was first introduced in L12B. Corresponding features are needed in EPC nodes to make use of this RAN feature. The new optional feature Operator Defined QCI provides finer granularity and/or flexibility when more than the predefined QoS profiles are needed. An operator QCI profile can be used for either GBR or non-gbr bearers. Up to 9 QoS profiles can be defined in addition to the 9 standard QoS profiles. 7.1 Feature Description The operator can define up to 9 QoS profiles in addition to the 9 standard QoS profiles, identified with QCI 1 through 9 Each operator defined QoS profile is identified with one QCI that can be chosen arbitrarily in the range For each operator defined QCI an instance of all per-qci operator parameters is created, as well as an instance of all per-qci counters. Possible to configure 9 Operator Defined QCI profiles besides existing Predefined/Standard profiles any value for QCI in range number of simultaneously used operator defined QCIs is limited to 9 (in addition to the predefined QCIs) Figure 6-26: Operator defined QCI -Feature Overview If operator defined QCI value is requested by the MME, enodeb should check that feature Operator Defined QCI is OPERABLE and QCI is defined in enb. If feature is set to INOPERABLE or QCI not defined in enodeb, the node will map the default QCI profile to affected bearer and mark the remap in separate PM event (from set up/modify procedure). This is also applicable during a handover. The feature provides more flexibility and higher level of diversity in tailoring different QoS policies to services and subscriber groups Ericsson AB 2012 LZT R1A

273 Capacity Management The subscriber context in the HSS needs to associate a QCI in the range with services for these QCIs to take effect The enodeb QCI table needs to be configured with such QCIs. 7.2 Configuration New MO class for Operator Defined QCI. The same attributes as class QciProfilePredefined, however, there is a new license handling class. These new options to have QCI profiles Added/removed/changed only take effect after a cell lock-unlock. After creating a new QCI profile instance in the enb, cell lock-unlock must be performed for the profile to be usable. After removing a QCI profile instance with OSS, cell lock-unlock must be performed for the profile to be removed and not usable anymore. Changes of the values of most parameters in an operator defined QCI profile take effect only after a cell lock-unlock is performed. However, for some parameters, changed values become usable at next bearer set up/modify. 8 Capacity Licenses Capacity licenses are used to control access to basic system functions and capacity. # Connected Users licensecapacityconnectedusers RRC LTE Figure 6-27 Capacity License Most LTE features are priced per connected users. The connected user feature sets the enodeb licensed user capacity and facilitates "pay as you grow" pricing scheme for SW features. SW license keys enable enodeb capacity in terms of maximum allowed simultaneous Connected Users. The active users are defined as connected terminals served by the enodeb, residing in the RRC Connected state (as defined in 3GPP). LZT R1A Ericsson AB

274 LTE L13 Radio Network Functionality Through the appropriate observability for the capacity licenses, the system helps the operator to optimize the usage of the SW licenses on a per enodeb basis. The maximum range for the amount of simultaneous Connected Users orderable is 100 UEs # Channel Bandwith licensecapacitychannelbandwidth 1.4, 3, 5, 10, 15, 20 MHz Δf=15kHz 180 khz frequency LTE Figure 6-28 Capacity License The capacity channel bandwidth license indicates the system bandwidth that can be 1.4, 3, 5, 10, 15 or 20 MHz. # HW Related CapacityOutputPower (W) DlBasebandCapacity (Mbps) UlBasebandCapacity (Mbps) DlPrbCapacity (PRB s) UlPrbCapacity (PRB s) LTE Figure 6-29: Capacity Licence - HW Related The hardware architecture allows for different site deployments. HW Capacity Licenses determines: Radio capacity given by combination of Radio Units (RU) and Digital Units (DU) that can be controlled by DL and UL Baseband capacity. Power supply units dimensioned for the site need controlled by CapacityOutputPower Ericsson AB 2012 LZT R1A

275 Capacity Management 9 Parameters 9.1 Parameters for admission control Parameter dlnongbrratio ulnongbrratio dltransnwbandwidth ultransnwbandwidth resourcetype Description The wanted downlink resource utilization ratio of Non-GBR bearers, for example for Mobile Broadband users, on transport network and cell levels in congested situations. The wanted uplink resource utilization ratio of Non-GBR bearers, for example for Mobile Broadband users, on transport network and cell levels in congested situations. Downlink transport network bandwidth for LTE. Uplink transport network bandwidth for LTE. Indicates if the resource type of the QoS Class Identifier (QCI) is Guaranteed Bit Rate (GBR) or non-gbr as defined in 3GPP TS paarpoverride The parameter is read-only and dependant on the QCI parameter setting. Allocation Retention Priority level received from the Core Network that enb interprets as the identifier for Privileged Access. The value zero will disable Privileged Access ARP override. 9.2 Affected Parameters for the Dynamic GBR Admission Control feature Parameter dlnongbrratio ulnongbrratio qciactuning Description Sets the downlink threshold for the Dynamic GBR Admission Control feature. Sets the uplink threshold for the Dynamic GBR Admission Control feature. Tuning factor per QCI for the basic admission control function. Needs the Dynamic GBR Admission Control feature to be activated to take effect. 9.3 Parameters introduced by the feature Operator Defined QCI. All parameters included in the new MO class QciProfileOperatorDefined are the same as the parameters included in the already existing MO class QciProfilePredefined Parameter abspriooverride aqmmode counteractivemode Table 1 Description Introduced Parameters Indicates if the data radio bearer is subject to Absolute Priority Override. Active Queue Management (AQM) mode Defines the interpretation of active for the QCI: FALSE : Active time is measured in the case of: DRB: as the time when there is data in a buffer, DL and/or UL, with an added guard period of 100 ms LZT R1A Ericsson AB

276 LTE L13 Radio Network Functionality datafwdperqcienabled dlminbitrate drxpriority drxprofileref dscp logicalchannelgroupref measreportconfigparams pdb thereafter UE: as the time when there is data in a buffer, DL and/or UL, with an added guard period of 100 ms thereafter in any of the DRBs connected to the UE The pmcounters for active E-RABs are stepped in case of data in a buffer, DL and/or UL. TRUE: Active time is measured in the case of: DRB: from the first data occurence in a buffer to the last data occurence in a buffer for that DRB UE: from the first data occurence in a buffer to the last data occurence in a buffer for any DRB associated with the UE The pmcounters for active E-RABs are always stepped since the DRB is considered always active Enables forwarding of data for this QCI. The scheduler will attempt to achieve minbitrate for all bearers before giving any user a higher rate. Value 0 means that the minrate scheduler is not used. The relative priority among the DRX profiles, i.e. if the bearer that is setup with this QCI has a higher DRX priority than any of the existing bearers, the DRX configuration will be set to those selected by the drxprofileid for this QCI. The drxpriority has to be unique across all the configured QciProfilePredefined and QciProfileOperatorDefined MOC instances except for instances where the drxprofileid is the same. That is, instances that share the same drxprofileid may have the same drxpriority value. Also note that larger drxpriority values indicate higher relative priority. Points out the DRX profile associated with this QCI. If not set the default reference will be to the DrxProfile instance corresponding to Local Distinguished Name (LDN) = ManagementElement=1, ENodeBFunction=1, DRxProfile=0. The Differentiated Services Code Point for a Quality of Service Class Indicator (QCI). This corresponds to mappings from RAN QoS to Transport Network QoS. Refers to an instance of LogicalChannelGroup. Assigns a Logical Channel Group to a Quality of Service Class Indicator (QCI). Encapsulates the offset quantities of the mobility measurement thresholds. Only absolute mobility measurement thresholds are considered. The structure holds, for example, offsets for the thresholds a1thresholdrsrpprim, a2thresholdrsrpprim, and a5threshold1rsrp. The offsets are meant mainly to have values that differentiate the measurement threhsolds between QCIs The contribution fromquality of Service Class Indicator (QCI)-related priority as defined in 3GPP TS (Release 8). enodeb to the Packet Delay Budget (PDB) for a QCI. Packet delays outside enodeb e.g. in the transport network are excluded. For more information about PDB refer to TS priority Quality of Service Class Indicator (QCI)-related priority as defined in 3GPP TS (Release 8). qci qciactuning qcisubscriptionquanta relativepriority resourceallocationstrategy resourcetype rlcmode rohcenabled schedulingalgorithm servicetype srsallocationstrategy ulminbitrate Quality of Service Class Indicator (QCI) as defined in 3GPP TS (Release 8). Each instance of this MO has a unique value for the QCI. Tuning factor per QCI for admission control. Only valid for QCIs with resource type GBR. Normalized subscription quantity associated with the specific the QCI. Specifies the subscription cost of a bearer with this predefined profile. The subscription cost is used for traffic load balancing purposes. The relative priority associated with a QCI. Defines the resource allocation strategy of the QoS Class Identifier (QCI). Indicates if the resource type of the Quality of Service Class Indicator (QCI) is Guaranteed Bit Rate (GBR) or non-gbr as defined in 3GPP TS RLC Mode Indicates if Robust Header Compression (ROHC) is enabled. Specifies which scheduling algorithm is to be used for a certain QCI. Indicates the service that the bearer is used for. If the parameter srsallocationstrategy in the QoS configuration for a Data radio Bearer assigned to a UE is set to ACTIVATED, then an attempt is made to allocate sounding for a UE. If several Data Radio Bearers are setup towards the UE with different QoS configurations, and the QoS configurations has different QCI parameter, an algorithm using the priority parameter in the QoS configuration, will resolve which QoS configuration that will define sounding The scheduler will attempt to achieve minbitrate for all bearers before giving any user a higher rate. Value 0 means that the minrate scheduler is not used. 9.4 Capacity Licenses MO Class Description CapacityChannelBandwidth Capacity for Channel Bandwidth Ericsson AB 2012 LZT R1A

277 Capacity Management CapacityConnectedUsers DlBasebandCapacity CapacityOutputPower UlBasebandCapacity Capacity for Connected Users Downlink Baseband Capacity Capacity for Output Power Uplink Baseband Capacity Grace period is an option for the following capacity licenses: CapacityNoActiveUsers DlBasebandCapacity UlBasebandCapacity When the RBS reaches the licensed capacity level installed it automatically adjusts to allow full capacity for a couple of days. Operators can then determine the required capacity. After these days the LTE RBS returns the capacity license value to the installed license value again. 9.5 ARP based Admission Control Parameter Description arpbasedpreemptionstate Activates or deactivates ARP-based UE preemption LZT R1A Ericsson AB

278 LTE L13 Radio Network Functionality 9.6 Differentiated Admission Control Parameter Description arpbasedpreemptionstate tinactivitytimer Activates or deactivates the Allocation and Retention Priority (ARP) based pre-emption function. The value of the attribute is irrelevant when no valid license key is installed for the Differentiated Admission Control feature. The time a UE can be inactive before it is released. This is legacy functionality. This feature now adds the following aspect: Let mintinactivitytimer be the smallest possible value of tinactivitytimer. Then UEs inactive for longer than mintinactivitytimer are eligible for Early UE Pre-emption. Consequently, if tinactivitytimer is set ot its smallest value there will be no UEs eligible for Early UE Pre-emption Ericsson AB 2012 LZT R1A

279 Automated Neighbor Relations 7 Mobility Objectives After this chapter the participants will be able to: Explain the purpose and function of Intra-LTE Handover, Inter-Radio Access Technologies (IRAT) Mobility and Inter Frequency and IRAT Session Continuity 1. Explain Intra LTE Handover 2. Explain Coverage Triggered Session Continuity 3. Describe the interworking with GRAN 4. Describe the interworking with UTRAN 5. Describe the interworking with CDMA Distinguish between release with redirect and handover 7. Detail what type of events trigger measurement reports to be sent to the enb 8. Describe the purpose of the handover evaluation algorithm and Best Cell Evaluation 9. Explain CS Fallback Figure 7-1: Objectives of Chapter 7 LZT R1A Ericsson AB

280 LTE L13 Radio Network Functionality 1 LTE Mobility The LTE mobility can be divided into Intra-LTE mobility and Inter-LTE mobility (inter-working with 2G/3G and CDMA 2000). RRC_CONNECTED mode mobility Intra-LTE Handover - within one MME pool Intra-eNode B Inter-eNode B Intra LTE Handover - Inter MME pool Inter-RAT Mobility Release with redirect RRC_IDLE mode mobility Cell reselection with TA update Figure 7-2 LTE Mobility Further we do distinguish between Intra LTE Mobility within one MME Pool and between MME pools. MME in Pool S-GW S-GW MME in Pool MME MME MME MME MME S-GW S-GW S1 HO X2 HO S1 CP S1 UP X2 CP and UP Figure 7-3 Intra MME Network Architecture Configuration In LTE, handovers are network controlled UE assisted handovers. As long as the UE moves between enbs that belong to the same pooling area where the UE is currently registered, the handovers are executed via the X2 interface. X2 interface can be set up via ANR or manually via OSS see Figure Ericsson AB 2012 LZT R1A

281 Automated Neighbor Relations MME S11 CP S-GW S1 CP S1 UP S1 CP S1 UP X2 UP X2 CP Automatic-ANR Manually via OSS Figure 7-4 Intra MME Handover Network Architecture Configuration In cases when the UE moves between enbs that belong to different pooling areas the handover procedure necessarily has to be executed via the S1 interface. In such cases at least the MME function, holding the UE context has to be relocated from one MME node in the first pool to another MME node in the second pool. The intra LTE connected mode mobility principles are summarized below: UE measures on serving cell and scans all neighboring intra-lte cells (504 PCIs) RS: UE measures RSRP & RSRQ RRC (measurement configuration): HO parameters (intra LTE carrier frequency, offsets, IRAT cell lists, Hyst etc) Measurement reports -> No UE neighbor list for intra-lte -> Detected good cells are reported -> IRAT cell lists are used RS: UE measures RSRP & RSRQ HO? Best Cell Evaluation enb makes HO decision based on UE measurements Serving cell Figure 7-5 Connected mode mobility Event? Mn HysteresisA 3 Ms a3offset Neighboring cell LZT R1A Ericsson AB

282 LTE L13 Radio Network Functionality 1.1 LTE Measurements The Intra-LTE Handover feature is based on measurements and coverage triggers evaluated by the user equipment. The user equipment measurements are reported to the serving RBS which makes the ultimate decision on inter-cell handover. RSRP Reference Signal Received Power RSRQ Reference Signal Received Quality RS Reference Signal RSRP RSRQ Figure 7-6: LTE Measurements Two types of measurements are used in the handover evaluation process: RSRP (Reference Signal Received Power) which represents the mean measured power per reference symbol RSRQ (Reference Signal Received Quality) which provides an indication of the reference signal quality The Intra-LTE handover can be set to trigger on the RSRP value or the RSRQ value and the measurement reports sent by the user equipment contain either or both of these values Measurement Configuration Measurement Configuration in Connected Mode The measurement configuration to be sent to the UE is defined in the RBS and for each cell (EUTRANCellFDD MO). For each EUTRANCellFDD there is one instance of MO UeMeasControl. UEMeasControl consist of three parts: Measurement configuration for Intra LTE MO ReportConfigEutraBestCell Measurement Configuration to support IRAT is selected by setting the parameter badcoveragemeasselection: MO ReportConfigEUtraBadCoveragePrim and MO ReportConfigEUtraBadCoverageSec Ericsson AB 2012 LZT R1A

283 Automated Neighbor Relations Poor coverage evaluation uses primary and secondary measurement parameters. This allows the UE to use different settings for two simultaneous measurements of type EventA2. Primary measurements use trigger quantity RSRP by default and secondary measurements use RSRQ by default. Both primary and secondary measurements can use the same trigger quantity Measurement Configuration initiated by Performance management using MO PmUeMeasControl and MO ReportConfigEUtraIntfaFreqPM Measurement configuration information is sent to the UE in RRCConnectionReconfiguration messages. For example, this may occur at RRC connection establishment or when the UE completes a handover to a LTE cell with different parameter settings. If the parameters are changed they do not update ongoing connections. RRC CONNECTION RECONFIGURATION (Measurement configuration) RRC CONNECTION RECONFIGURATION COMPLETE Measurement objects (measobjecttoaddmodifylist, measobjecttoremovelist) The objects which the UE shall perform the measurement on e.g. a carrier frequency or a list of neighbouring cell offsets or IRAT neighbouring cells. Reporting configurations (reportconfigtoaddmodifylist, reportconfigtoremovelist) Reporting criterion: periodical or event-triggered reporting Reporting format: quantities (e.g. number of cells to report) Measurement identity (measidtoaddmodifylist, measidtoremovelist) List of measurement identities, each identity links one measurement object with one reporting configuration. This is the reference number in the measurement report. Quantity configurations (quantityconfig) The quantity the UE shall measure as well as the associated firing parameters. E.g. RSRP (Reference Signal Received Power), RSRQ (Reference Signal Received Quality). One quantity for intra freq, one for inter and one for each RAT type Figure 7-7 Measurement Configuration in UE The UE measurements necessary for the handover are configured from the source enb via RRC. It is generally assumed that downlink measurements, done by the UE, are used for the handover decision. The reporting can be done on an event triggered and/or periodic basis. However, the UE is also required to detect and measure neighbor cells itself, since no neighbor cell list is configured in the measurement command for intra- LTE mobility. The measurement configurations that need to be indicated for inter-system measurements depend on the target system type and may include neighbor cell lists LZT R1A Ericsson AB

284 LTE L13 Radio Network Functionality PERIODICAL Periodical reporting. EVENT_A1 Serving cell becomes better than absolute threshold. EVENT_A2 Serving cell becomes worse than absolute threshold. EVENT_A3 Neighbor cell becomes amount of offset better than serving. EVENT_A4 Neighbor cell becomes better than absolute threshold. EVENT_A5 Serving cell becomes worse than absolute threshold1 AND neighbor cell becomes better than another absolute threshold2. Figure 7-8: Measurement Report: LTE Events (Intra and Inter frequency) The UE finds intra frequency cells and detects the identities by their synchronization signals. The UE performs handover measurements on detected cells reference symbols, filters and evaluates measurement results with respect to event conditions. If condition for start event is met a measurement report to source RBS is sent. If parameter TimeToTrigger is set the event is considered not to be fulfilled until the event condition is fulfilled during the TimeToTrigger. Then the event report is sent. Periodic reporting results in measurement reports being sent periodically as long as condition for stop event is not met. PERIODICAL Periodical reporting. EVENT_B1 IRAT neighbour becomes better than threshold EVENT_B2 Serving becomes worse than threshold1 and IRAT neighbour becomes better than threshold2 Figure 7-9: Measurement Report: LTE Events (IRAT) Ericsson AB 2012 LZT R1A

285 Automated Neighbor Relations UE measures on cells and reports only when it according to event criteria is met + UE transmit reports immediately and only when Network configured event criteria are fulfilled. Typical events: - A neighbour cell becomes offset better than serving cell (A3) - Serving cell becomes worse than an absolute threshold (A2) -... Figure 7-10 Event Triggered Measurement Reporting RRC: Measurement Report Reporting criteria Reporting threshold Hysteresis Time-to-trigger Reporting interval Measured results Event id Cell identity Measured measurement quantity Figure 7-11 RRC Measurement Reporting LZT R1A Ericsson AB

286 LTE L13 Radio Network Functionality 2 Intra LTE Handover The Intra-LTE Handover feature configures EventA3 as defined in 3GPP TS The EventA3 implies that one or several neighbor cells become better than the serving cell also when some offset and hysteresis values are taken into account "neighbor becomes offset better than serving". The process employed by the user equipment for the evaluation of surrounding cells uses parameters sent by the serving RBS to the user equipment. These parameters include hysteresis and offset values, time to trigger, and optionally cell individual offset margins. The Intra-LTE Handover feature is based on the evaluations reported to the RBS by the user equipment. The serving RBS uses the reports to select and prepare the target RBS, then ultimately conducts conclusion of the handover. User equipments use two alternative types of measurements in the cell evaluation process: Reference Signal Received Power (RSRP) representing the mean measured power per reference signal Reference Signal Received Quality (RSRQ) providing an indication of the reference signal quality The intra-lte handover can be set to trigger on the RSRP value or the RSRQ value. The measurement reports sent by the user equipment contain either or both of these values. There are three types of mobility procedures for Intra-LTE handover; intra-rbs handover, the X2 based inter RBS handover, and the S1 based inter RBS handover. The handover procedures are described in the following: The intra-rbs handover procedure is used when both the source and target cells reside in the same RBS. The X2 inter-rbs handover is primarily used when an X2 relation exists between source and target RBS. Both source and target RBS must be connected to the same MME. The S1 inter-rbs handover is primarily used when no X2 relation exists between source and target. Source and target RBS can be connected to the same or different MME. X2 and S1 handover procedures differ mainly in the signaling towards the core network. Both of them support packet forwarding for RLC AM as well as RLC UM from L13A. Packet forwarding reduces the interruption time at handovers by forwarding the buffered data from source to target enodeb Ericsson AB 2012 LZT R1A

287 Automated Neighbor Relations 2.1 Event A3 The user equipment uses either RSRP or RSRQ measurements to determine whether to enter the EventA3 condition. The triggerquantitya3 parameter is used to configure whether RSRP or RSRQ values are used to trigger EventA3. Measurements of RSRP and RSRQ are performed on the serving and detected neighboring cells. The measured values of RSRP and RSRQ can then be filtered based upon the settings of the filtercoefficienteutrarsrp and filtercoefficienteutrarsrq parameters. The filter averages a number of measurements in order to filter out the impact of large scale fast fading. RSRP / RSRQ Cell A To activate the function bestcellreleaseactive = TRUE triggerquantitya3 = trigger on RSRP or RSRQ UE measures neighbouring cells, measurement reports can be RSRP and/or RSRQ (reportquantitya3) smeasure (triggers on RSRP only) A3offset HysteresisA3 Enter Event A3 Firing can be applied to measurements before Reports are sent to the RBS timetotriggera3 Leave Event A3 Measurement Reports reportintervala3 Cell B reportamounta3 (0 = continual during event) Mn + cellindividualoffseteutran - hysteresisa3 > Ms + a3offset Time Figure 7-12: Event A3: Neighbor becomes amount offset better than serving The user equipment then uses an offset value, a3offset, and a hysteresis value, hysteresisa3, to determine whether to trigger the EventA3. Non default offset relationships use the value cellindividualoffseteutran instead of a3offset for the particular cell relationship. The expression used by the user equipment for evaluating entry to EventA3 is shown below: Mn + cellindividualoffseteutran - hysteresisa3 > Ms + a3offset Mn = measured value of the neighboring cell (either RSRP or RSRQ) Ms = measured value of the serving cell (either RSRP or RSRQ) LZT R1A Ericsson AB

288 LTE L13 Radio Network Functionality Once EventA3 is triggered, the user equipment will wait a predetermined time (timetotriggera3) before it commences sending measurement reports to the serving RBS. These measurement reports contain measurements for the serving cells and up to maxreportcellsa3 detected intra frequency neighbor cells. The reportquantitya3 parameter indicates whether RSRP or RSRQ measurements, or both, are to be included in the measurement reports. Measurement reports are sent periodically whilst the EventA3 condition is active. There are two parameters that control reporting: reportintervala3 determines the time interval between measurement reports. reportamounta3 indicates how many reports to send; a value of 0 indicates that the reports should be sent indefinitely whilst the EventA3 condition is active. The user equipment uses the same offset and hysteresis values to determine when to leave EventA3 when the serving cell improves in RSRP or RSRQ relative to the neighboring cells. The expression used by the UE is shown below: Mn + cellindividualoffseteutran + hysteresisa3 < Ms + a3offset Mn = measured value of the neighboring cell (either RSRP or RSRQ) Ms = measured value of the serving cell (either RSRP or RSRQ) The Measurement Reports will be Event triggered and resent periodically as long as the event is fulfilled Best Cell Evaluation Best cell evaluation in the enb makes it possible to: Set handover threshold relative to source cell Handle both intra-mme and inter-mme cells in evaluation Set cell specific offsets relative to handover threshold to compensate for cell DL/UL imbalance or to offset handover region. If best cell handover proposal is rejected and another cell that also have higher quality than the source cell is available that cell is proposed for handover. A neighbor list is configured in the enb that contains cell-id, Inter/Intra-MME relation, cell specific offset, etc Ericsson AB 2012 LZT R1A

289 Automated Neighbor Relations Best Cell Evaluation keeps a list with best cell candidates. The list is updated based on measurement reports. The size of the list should correspond to the max number of cell measurements that UE can report in a message report. The Best cell candidate list can contain both Intra and Inter MME LTE cells. A proposal for handover is triggered when current source cell no longer is best cell. If Best Cell Evaluation receives a measurement report while it already processes a measurement report the new report will be buffered. When starting a UE best cell evaluation the latest buffered measurement report will be considered. There is an operator controlled parameter minbestcellhoattempts that decides how many handover attempts shall be performed on a cell before the next best cell is chosen. ANR Measurement Report A3 Cell with no of HOattempts < minbestcellhoattempts exists Cell Unknown Chose best cell that fullfills noofhoattempts < minbestcellhoattempts noofhoattempts +1 Chose best cell Cell Known No NR Best cell X2 allowed X2 HO minbestcellhoattempts Figure 7-13 Best Cell Evaluation X2 not allowed S1 Handover The system information blocks SIB1, SIB3, and SIB4 conveys the information necessary for the UE to perform intra LTE mobility functions. 2.2 Inter Frequency Mobility The Inter Frequency Mobility consists of Coverage Triggered Inter Frequency Handover and Coverage Triggered Inter Frequency Session Continuity. The Coverage-Triggered Inter-Frequency Handover/Session Continuity operation is summarized below: LZT R1A Ericsson AB

290 LTE L13 Radio Network Functionality Licence featurestateinterfrequencysessioncontinuity featurestateinterfrequencyhandover Redirection info redirect prio (connectedmodemobilityprio) arfcnvalueeutrandl Cell reselection info SIB 5 (LTE IF) Event A2 Bad coverage detected NO RBS determines a set of candidate frequencies Is there an LTE frequency cell that fully covers the source cell? YES FALSE uemeasurementactive=? TRUE Event A1 Good coverage detected Session continuity a5b2mobilitytimer Event A5 Serving worse than threshold1 AND neighbor better than threshold2 Blind release with redirect (to one of the candidate freq) Release with redirect (to freq reported by A5) EUtranCellRelation ishoallowed=true REDIRECT mobilityaction=? HANDOVER covtriggeredblindhoallowed=true mobilityaction=handover coverageindicator=covers ishoallowed=true Handover X2 or S1 IF HO Blind HO Figure 7-14: Inter-Frequency Mobility The Source RBS configures an Event A2 (serving cell becomes worse than threshold) measurement in the UE. When the Event A2 measurement report is received from the UE by the serving RBS, the RBS determines the set of LTE candidate frequencies where the UE can be transferred. If any of the LTE candidate frequencies is an LTE frequency, the RBS checks if it has information about a cell on this frequency which is fully covering the source cell, and if so selects this cell as the target for (blind) handover. If none of the LTE frequency candidates has a cell which covers the UE, and the RBS is configured to start measurements in the case of a Event A2 measurement report, an Event A5 measurement (serving becomes worse than threshold1 and neighbor becomes better than threshold2) for each of the candidate frequencies will be configured in the UE. When an Event A5 measurement report is received from the UE, the RBS will investigate the reported cells and select a target cell which is possible to do a handover to. If none of the reported cells are possible to use for handover, the UE will be released with redirect to the frequency reported bye the Event A5 measurement. If the RBS is configured to not start measurements in the case of an Event A2 measurement report, no target cell will be selected and instead the UE will be released with redirect to one of the candidate frequencies Ericsson AB 2012 LZT R1A

291 Automated Neighbor Relations If a target cell was selected for handover, the RBS proceeds to prepare and execute the handover attempt. This is done in the same way as for Intra-LTE Handover, including the decision on whether to use an X2 or S1 connection when communicating with the RBS handling the target cell, in the case when the target cell is handled by a different RBS than the RBS handling the source cell. The events sent from the RBS during handover preparation and execution will contain information saying that it is an "inter-frequency handover" and whether the handover was initiated based on a pre-configured target cell (blind handover), or if an Event A5 measurement was used to find a target cell (measurement based handover). When the RBS has selected an LTE frequency as a candidate target for transferring the UE, the RBS will check if the UE should be transferred there with a blind release with redirect, a blind handover, or if an Event A5 measurement should be started in the UE. To indicate that blind handover should be used for this target frequency, the operator needs to configure the RBS in the following way: The attribute covtriggerdblindhoallowed for the serving cell must be set to the value true. The attribute mobilityaction must have the value HANDOVER for the EUtranFreqRelation MO representing the target frequency under the source cell in the MOM. There has to exist a EUtranCellRelation below the target EUtranFreqRelation which indicates a cell which is covering the serving cell completely, and therefore is safe to send the UE to from the serving cell without measuring. This is achieved in the EUtranCellRelation by setting the attribute coverageindicator to the value COVERS. Also, the attribute ishoallowed must have the value true. This cell relation represents the target cell in the resulting blind handover. Note that it is not required that the attribute uemeasurementactive has the value false. Only if there are no cell relations that can be used for blind handover will the value of uemeasurementactive be used to determine if a blind release with redirect should be triggered, or if an Event A5 measurement should be started in the UE. When the RBS has selected an LTE frequency as a candidate target for transferring the UE, the RBS will check if the UE should be transferred there with a blind release with redirect, a blind handover, or if an Event A5 measurement should be started in the UE. If the decision is to start an Event A5 measurement, the RBS will then wait for the resulting report and then decide if the UE should be transferred to the target frequency with a release with redirect, or a handover to one of the cells included in the measurement report. To indicate that measurement based handover should be used for this target frequency, the operator needs to configure the RBS in the following way: LZT R1A Ericsson AB

292 LTE L13 Radio Network Functionality The attribute uemeasurementactive must be set to the value true. The attribute mobilityaction must have the value HANDOVER for the EUtranFreqRelation MO representing the target frequency under the source cell in the MOM. For every cell that might be reported by the Event A5 measurement and that should be possible to use as target cells, there has to exist a EUtranCellRelation below the target EUtranFreqRelation where the attribute ishoallowed has the value true. Note that if the received Event A5 measurement report only contains cells which either does not exist in the MOM, or for which there is no cell relation from the source cell, or for which the attribute ishoallowed has the value false, the RBS will transfer the UE to the target frequency using a release with redirect Event A2 Event A2 follows a similar principle as event A3. Having two measurement thresholds for poor coverage evaluation allows for the use of different hysteresis and timer values for the different threshold values. The process for evaluation using the secondary parameters and threshold is identical; however, different parameter values can be used. As an example, the triggerquantitya2prim can be set to trigger on RSRP measurements with a hysteresisa2prim value of 1 db and the triggerquantitya2sec can be set to trigger on RSRQ measurements with a hysteresisa2sec value of 2 db. The picture shows how the parameters are used by the UE for poor coverage evaluation using the primary threshold: Ericsson AB 2012 LZT R1A

293 Automated Neighbor Relations RSRP / RSRQ Cell A triggerquantitya2prim = trigger on RSRP or RSRQ UE measures neighboring cells measurement reports can be RSRP and/or RSRQ (reportquantitya2prim) smeasure (triggers on RSRP only) hysteresisa2prim a2thresholdrsrpprim or a2thresholdrsrqprim Enter Event A2 Leave Event A2 Measurement Report Filtering can be applied to measurements before Reports are sent to the RBS timetotriggera2prim reportamounta2 (0 = continual during event) Ms+hysteresisA2Prim < a2thresholdprimary Time Figure 7-15: Event A2: Serving becomes worse than absolute threshold The UE uses parameters sent by the RBS to determine when to perform measurements. Measurements begin on the serving and neighboring cells when the RSRP of the serving cell falls below the value defined in the smeasure parameter. This parameter is also the trigger value for measurements used in intra-lte handover. The evaluation process for the primary trigger is described below. The process for the secondary trigger is identical and occurs in parallel with the primary process if the parameter badcoveragemeasselection is set to a value of BOTH. The triggerquantitya2prim parameter determines whether RSRP or RSRQ values trigger Event A2 for the primary criteria. If the primary trigger is set to RSRP, then the threshold value set in a2thresholdrsrpprim is sent to the UE, otherwise the RSRQ threshold value set in a2thresholdrsrqprim is sent. The measured values of RSRP and RSRQ are filtered based on settings of the filtercoefficienteutrarsrp and filtercoefficienteutrarsrq parameters. The filter averages a number of measurements to remove the impact of fast fading. These filter values are also used by the UE in the intra-lte handover measurement process. The UE applies a hysteresis value, hysteresisa2prim, to determine whether to trigger the Event A2, as shown above. LZT R1A Ericsson AB

294 LTE L13 Radio Network Functionality The expression used by the UE for evaluating entry to Event A2 is shown in the following expression: Ms + HysteresisA2Prim < a2thresholdprimary Expression 1: Condition for Entering EventA2 where: Ms = measured value of the serving cell (either RSRP or RSRQ) a2thresholdprimary is the value of either a2thresholdrsrpprim or a2thresholdrsrqprim depending on whether the trigger quantity is RSRP or RSRQ. When Event A2 is triggered, the UE waits a predetermined time (timetotriggera2prim) before it begins sending measurement reports to the serving RBS. The reports contain measurements for the serving cell. The reportquantitya2prim parameter indicates whether RSRP, or RSRQ measurements, or both, are to be included in the measurement reports. Measurement reports are sent periodically. There are two parameters that control reporting: reportintervala2prim determines the time interval between measurement reports. reportamounta2prim indicates the number of reports to send. The value 0 indicates that the reports are sent indefinitely while the Event A2 condition is active. The UE uses the same hysteresis value to determine when to leave Event A2 once the serving cell improves in RSRP or RSRQ relative to the neighboring cells. The expression used by the UE is shown in the following equation: Ms - HysteresisA2Prim > a2thresholdprimary Expression2: Condition for Leaving EventA2 where: Ms = measured value of the serving cell, either RSRP or RSRQ a2thresholdprimary is the value of either a2thresholdrsrpprim or a2thresholdrsrqprim depending on whether the trigger quantity is RSRP or RSRQ. The Measurement Reports are event-triggered and are resent periodically as long as the event is fulfilled Ericsson AB 2012 LZT R1A

295 Automated Neighbor Relations Event A5 Event A5 is used by Inter Frequency Handover in order to find a frequency to redirect to or a cell to make handover to. To activate the function uemeasurementactive = TRUE RSRP / RSRQ Cell A triggerquantitya5 = trigger on RSRP (or RSRQ) UE measure neighboring cells measurement smeasure (triggers on RSRP only) a5threshold2rsrp or a5threshold2rsrq hysteresisa5 a5threshold1rsrp or a5threshold1rsrq Cell B Enter Event A5 Leave Event A5 timetotriggera5 Measurement Report Both criteria must be fulfilled in order to enter A5 Ms Hys Thresh 1 Mn Hys Thresh 2 Time Mn Hys Thresh 2 Only one criterium must be fulfilled in order to leave A5 Figure 7-16: Event A5: Serving worse than threshold 1 AND Neighbor better than threshold 2 LZT R1A Ericsson AB

296 LTE L13 Radio Network Functionality Event A1 The redirection to another frequency (or RAT as we will see later) may be cancelled by event A1 ( Good Coverage ). RSRP / RSRQ Cell A triggerquantitya2prim = trigger on RSRP or RSRQ (follows A2) UE measure neighboring cells measurement reports can be RSRP or RSRQ hysteresisa1prim a1thresholdrsrpprim or a1thresholdrsrqprim Filtering can be applied to measurements before Reports are sent to the RBS Enter Event A1 timetotriggera1prim Measurement Report Leave Event A1 Ms-hysteresisA2Prim < a2thresholdprimary Time Figure 7-17: Event A1: Serving becomes better than absolute threshold 3 IRAT Mobility This chapter deals with inter-working between LTE (E-UTRAN) and UTRAN GERAN CDMA2000 There are in principle two ways of inter-working between LTE and other RATs. The inter-working can be performed by a prepared handover (network controlled) where the UE does not leave the CONNECTED state (Handover) or by a cell reselection (UE controlled or network assisted) where UE via IDLE state performs network assisted cell reselection, see below Ericsson AB 2012 LZT R1A

297 Automated Neighbor Relations Release with Redirect (Coverage Triggered Session Continuity) Handover RRC-CONNECTED RRC-IDLE Bad coverage detection triggers Release with Redirect: Redirect Information Frequency RRC-CONNECTED Handover Command Move to reserved resources RRC-CONNECTED CellReselection according to redirect information (GSM, WCDMA, ehrpd, LTE IF) RRC-CONNECTED Figure 7-18 IRAT Mobility The following features are associated with IRAT Mobility: Coverage Triggered WCDMA IRAT Handover, Coverage Triggered WCDMA Session Continuity, Coverage Triggered GERAN Session Continuity, Coverage Triggered CDMA-eHRPD Session Continuity. SRVCC Handover to UTRAN CS Fallback 3.1 Coverage triggered Session Continuity The Coverage-Triggered WCDMA Session Continuity operation is summarized in the next figure: LZT R1A Ericsson AB

298 LTE L13 Radio Network Functionality Licence featurestatewcdmasessioncontinuity featurestategsmsessioncontinuity featurestatecdma200sessioncontinuity Redirection info arfcnvalueutrandl arfcnvaluegerandl hrpdbandclass, freqcdma Cell reselection info SIB 6 (WCDMA) SIB 7 (GERAN) SIB 8 (CDMA) Event A2 Bad coverage detected Event B2 Serving worse than threshold1 AND IRAT better than threshold2 Event A1 Good coverage detected TRUE RBS determines a set of candidate frequencies uemeasurementactive FALSE Release with redirect (to indicated IRAT freq) a5b2mobilitytimer Blind release with redirect (to one of the candidate freq) Figure 7-19: Coverage Triggered IRAT Session Continuity The RBS determines a set of frequencies, LTE or IRAT, where the UE can be transferred when it encounters poor coverage in the current cell. This is done as follows: First find all frequency relations, that is, all instances of EUtranFreqRelation, GeranFreqGroupRelation, UtranFreqRelation, and Cdma2000FreqBandRelation, for the current cell as follows: o o The corresponding Coverage-Triggered Session Continuity feature is active and enabled. For example, only if the Coverage- Triggered WCDMA Session Continuity feature is active and enabled will any UtranFreqRelation be included. The UE has capabilities that enable it to operate on this frequency. From this set, select the frequency relations with the highest value of the connectedmodemobilityprio attribute (it is possible that several frequency relation MOs share the highest value for this attribute and in this case they are all selected). If the set of candidate frequencies found is non-empty, the UE is configured to send an event-triggered measurement report, Event A2, when coverage provided by the serving cell drops below the threshold for poor coverage of a parameterdefined margin. After some time, the Event A2 measurement report is received from the UE Ericsson AB 2012 LZT R1A

299 Automated Neighbor Relations The RBS will first check if a blind handover is possible, that is a handover to a pre-defined cell. This is enabled by a separate feature, Coverage-Triggered Inter- Frequency Handover, and is not further described here, except that only cells using a candidate frequency found in Step 2 Page 7 would be considered. Measurements should be started if the uemeasurementsactive attribute is true and the candidate frequencies selected in earlier contains at least one frequency which can be measured. A frequency can be measured if: The UE supports measurements on the frequency. The RBS supports measurements on the frequency: o o The RBS supports measurements on all LTE and GERAN frequencies. The RBS supports measurements on WCDMA and CDMAeHRPD only if there are configured cells for these frequencies, that is, MO instances of UtranCellRelation and Cdma2000CellRelation If measurements should not be started, initiate a release with redirect to one of the selected frequencies. Since these are all on the same priority level, the RBS can pick any one of them. It is unspecified which one the RBS will pick, but given the same selection of frequencies, the RBS will always pick the same frequency. If measurements should be started, start "Target good enough" measurements (Event A5 for LTE frequencies and Event B2 for IRAT frequencies) to detect the candidate frequencies selected earlier which can be measured. Also start Event A1 measurements to detect if the UE moves into good coverage again. At the same time start the timer a5b2mobilitytimer. An Event A1 Good Coverage measurement report arrives from the UE. The RBS will stop the Event A5/B2 and A1 measurements in the UE, and wait for another Event A2 measurement report. If a5b2mobilitytimer, times out the RBS will release the UE with a redirection to one of the frequencies for which Event A5/B2 measurements were started. An Event A5/B2 measurement report arrives from the UE for one of the frequencies. The RBS will first check if a handover is possible to one of the reported cells. This is enabled by a separate Coverage-Triggered Inter-Frequency Handover feature. Initiate a release with redirect to the frequency of the measurement report. LZT R1A Ericsson AB

300 LTE L13 Radio Network Functionality The poor coverage evaluation can use primary and secondary measurement parameters to allow the UE to use different settings for simultaneous measurements for Event A2. The primary measurements use RSRP by default and the secondary measurements use RSRQ by defaults. Both primary and secondary measurements can use the same measurement quantity. Measurement configuration information is sent to the UE in RRCConnectionReconfiguration messages. A new version of this message is sent to the UE whenever there is an update of the parameters. For example, this may occur when the UE completes a handover to a LTE cell with different parameter settings. The poor coverage evaluation parameters of the RRCConnectionReconfiguration are contained in the MO UeMeasControl, a child of the MO EUtranCellFDD Event B2 Event B2 is used by IRAT Coverage Triggered Session Continuity in order to decide which type of release (blind or not) will be performed. To activate the function uemeasurementactive = TRUE RSRP / RSRQ Cell A triggerquantityb2 = trigger on RSRP UE measures neighbouring cells measurement smeasure (triggers on RSRP only) b2threshold2ecn0utra or b2threshold2rscputra or b2thresholdgeran or b2thresholdcdma2000 timetotriggerb2 Enter Event B2 hysteresisb2 Measurement Report Cell B Leave Event B2 b2threshold1rsrp or b2threshold1rsrq Time Both criteria must be fulfilled in order to enter B2 Ms Hys Thresh 1 Mn Hys Thresh 2 Mn Hys Thresh 2 Only one criterium must be fulfilled in order to leave B2 Figure 7-20: Event B2: Serving worse than threshold 1 AND IRAT Neighbor better than threshold Ericsson AB 2012 LZT R1A

301 Automated Neighbor Relations 3.2 Coverage triggered WCDMA IRAT Handover Event A2 Bad coverage detected The prerequisite for this feature is the activation and proper configuration of the coverage triggered WCDMA Session Continuity feature. The Coverage-Triggered WCDMA IRAT Handover/Session Continuity operation is summarized in the next figure: Licence featurestatewcdmahandover featurestategsmsessioncontinuity featurestatecdma200sessioncontinuity featurestatewcdmasessioncontinuity Redirection info redirect prio (connectedmodemobilityprio or SPID)) arfcnvalueutrandl arfcnvaluegerandl hrpdbandclass, freqcdma Cell reselection info SIB 6 (Utran) SIB 7 (GERAN) SIB 8 (CDMA) *also taking into account possible SPID values NO RBS determines a set of candidate Is there a WCDMA frequencies* frequency cell that fully covers the source cell? YES FALSE uemeasurementactive=? a5b2mobilitytimer covtriggeredblindhoallowed=true mobilityaction=handover coverageindicator=covers ishoallowed=true TRUE Event A1 Good coverage detected Session continuity Event B2 Serving worse than threshold1 AND IRAT neighbor better than threshold2 Handover Blind release with redirect (to one of the candidate freq) Release with redirect (to freq reported by B2) mobilityaction=? UtranCellRelation ishoallowed=true externalutrancellfdd lac 0 and rac -1 IRAT HO REDIRECT HANDOVER Blind IRAT HO Figure 7-21: IRAT Mobility (WCDMA) The source enodeb configures an Event A2 (serving cell becomes worse than threshold) measurement in the UE. When the UE is in poor coverage area, it sends a measurement report with event A2 to the enodeb. The enodeb determines the set of candidate frequencies where the UE can be transferred. 1. If any of the candidate frequencies is a WCDMA frequency with the attribute mobilityaction set to HANDOVER and a cell on this frequency is not a blacklisted cell (ishoallowed=true) and is fully covering the source LTE cell (coverageindicator =COVERS), the enodeb triggers a WCDMA blind handover to this cell. 2. If none of the WCDMA frequency candidates has a cell that fully covers the source cell, and the enodeb is configured to start measurements when LZT R1A Ericsson AB

302 LTE L13 Radio Network Functionality receiving an Event A2 measurement report, an Event B2 measurement for each of the candidate WCDMA frequencies will be configured in the UE. When an Event B2 measurement report is received from the UE, the enodeb will investigate the reported cells and select a target cell to which is possible to do a handover. If none of the reported cells can be used for handover, the UE will be released with redirect, as described in WCDMA Session Continuity, Coverage-Triggered, to the frequency reported by the Event B2 measurement. If the enodeb is configured to not start measurements for an Event A2 measurement report or the enodeb times out (A2B5MobilityTimer) while waiting for the B2 measurement report, no target cell will be selected and instead the UE will be released with redirect to one of the candidate frequencies Configure for Outgoing WCDMA Handover Regardless of whether the enodeb executes a blind or measurement-based handover for a certain UE, for a certain WCDMA cell to be usable as target cell for the handover, the following two parameters must be set to values that are different from the default value: lac Location area code: Must be sent to the MME to correctly identify the target SGSN that serves the target WCDMA RNC. rac Routing area code: Must be sent to the MME to uniquely be able to identify he target SGSN in the target location area of the target WCDMA cell. For a WCDMA cell ExternalUtranCellFDD Managed Object (MO) instance that is manually created by an operator, it is expected that the values for the rac and lac parameters are filled in at creation of the MO instance. This is not mandated, but very much recommended Configure for Blind WCDMA Handover When the RBS has selected a WCDMA frequency as a candidate target for transferring the UE, the RBS will check whether the UE should be transferred there with a blind release with redirect, a blind handover, or whether an Event B2 measurement should be started in the UE. To indicate that blind handover should be used for this target frequency, the operator must configure the RBS in the following ways: The covtriggerdblindhoallowed attribute for the serving cell must be set to the value true Ericsson AB 2012 LZT R1A

303 Automated Neighbor Relations The mobilityaction attribute must have the value HANDOVER for the UtranFreqRelation MO representing the target frequency under the source cell in the Managed Object Model (MOM). There must be an UtranCellRelation below the UtranFreqRelation target that indicates a cell that is covering the serving cell completely, and therefore is safe to send the UE to from the serving cell without measuring. This is achieved in the UtranCellRelation by setting he coverageindicator attribute to the value COVERS. Also, the shoallowed attribute must have the value true. This cell relation represents the target cell in the resulting blind handover. Note: It is not required that the uemeasurementactive attribute has the value false. Only if there are no cell relations that can be used for blind handover will the value of uemeasurementactive be used to determine whether a blind release with redirect should be triggered, or whether an Event B2 measurement should be started in the UE. The coverageindicator parameter has the values NONE, COVERS, OVERLAP, CONTAINED_IN. These are illustrated in the following figure. COVERS CONTAINED_IN OVERLAP WCDMA LTE LTE LTE WCDMA WCDMA NONE LTE WCDMA Figure 7-22: coverageindicator values Configure for Measurement-Based WCDMA Handover When the RBS has selected a WCDMA frequency as a candidate target for transferring the UE, the RBS will check whether the UE should be transferred here with a blind release with redirect, a blind handover, or whether an Event B2 measurement should be started in the UE. If the decision is to start an Event B2 measurement, the RBS will then wait for the resulting report and then decide whether the UE should be transferred to the target frequency with a release with redirect, or a handover to one of the cells included in the measurement report. LZT R1A Ericsson AB

304 LTE L13 Radio Network Functionality To indicate that measurement-based handover should be used for this target frequency, the operator must configure the RBS in the following ways: The uemeasurementactive attribute must be set to the value true. The mobilityaction attribute must have the value HANDOVER for the UtranFreqRelation MO representing the target WCDMA frequency under the source cell in the MOM. For every cell that might be reported by the Event B2 measurement and that should be possible to use as target cells, there must be an UtranCellRelation below the UtranFreqRelation target where the ishoallowed attribute has the value true. Note: If the received Event B2 measurement report only contains cells that either do not exist in the MOM, for which there is no cell relation from the source cell, or for which the ishoallowed attribute has the value false, the RBS will transfer the UE to the target frequency using a release with redirect. 3.3 SRVCC Handover to UTRAN When QCI 1 bearers are used, it is assumed that there is a VoIP service ongoing. Then, in case of bad coverage, a B2 event may trigger SRVCC (Single Radio Voice Call Continuity) handover to UTRAN, with the voice bearers being handed over to the CS domain. SRVCC Single Radio Voice Call Continuity - Handover of VoIP call in LTE (VoLTE) to CS call in WCDMA - PS to CS transition using voice IRAT handover - Done with SRVCC procedures, triggered by A2 and B2 events - Triggered by enodeb, executed by Core NW and IMS VoLTE WCDMA CS RBS RBS Figure 7-23: Background SRVCC Handover to UTRAN What is it? Event B2 can be configured when A2 search event is fulfilled, but blind SRVCC handover to UTRAN can also be triggered directly from A2 bad coverage event (if A2 search is not used) Ericsson AB 2012 LZT R1A

305 Automated Neighbor Relations Basically, the SRVCC handover to UTRAN is triggered by the same functions as ordinary IRAT handover to UTRAN. Also, voiceprio can be used to set priorities on frequencies for UEs with QCI 1 bearer(s) which enables a different frequency plan for voice. VoIP call transition to the CS domain in a WCDMA cell when poor coverage is detected in the serving LTE cell Enhanced control over how voice bearers are transferred between frequencies and cells Target node (RNC) can use different admission criteria than would be used at a plain connection setup This makes it possible to improve the probability that a VoIP service is not interrupted in a network with high load. Figure 7-24: SRVCC benefits Licence featurestatesrvcctoutran featurestatewcdmasessioncontinuity Redirection info redirect prio (connectedmodemobilityprio/voiceprio or SPID) Do nothing (stay in source cell) Event A2 Bad coverage detected coverageindicator=none NO RBS determines a set of candidate frequencies Is there a WCDMA cell that fully covers the source cell and is SRVCC capable? YES Event A1 Good coverage detected (both RSRP&RSRQ) a5b2mobilitytimer Event B2 srvcccapability NOT_SUPPORTED UtranCellRelation ishoallowed=true externalutrancellfdd lac 0 and rac -1 Measurement based SRVCC HO srvcccapability NOT_SUPPORTED covtriggeredblindhoallowed=true coverageindicator=covers ishoallowed=true Blind SRVCC HO Figure 7-25: SRVCC HO to UTRAN - UE has at least one QCI1 bearer The activation of this feature brings some general voice improvement for every voice call using QCI=1 as follows: Never Release with Re-direct o Release with Re-direct will actively drop the voice call Try to keep the voice call alive as far as possible LZT R1A Ericsson AB

306 LTE L13 Radio Network Functionality voiceprio parameter to be used for frequency priority o Instead of connectedmodemobilityprio for voice connections o Makes it possible to steer voice traffic separately from other traffic o Available in MOM on Utran Cell, EUtran Cell and in SPID structure SRVCC may co-exist with CS Fallback (see later this chapter) when there is no VoIP support in UE, LTE or Core/IMS network. Then the UE initiates CS Fallback (a MO call with an extended service request to the MME, as a CS Fallback indicator) before call is established. If the UE has both Voice and data PS bearers, the target cell must be CS + PS capable and there must be an operable feature for IRAT HO to UTRAN in order to transfer also the PS bearer to WCDMA. 3.4 Subscriber triggered Mobility Summary This feature enables the Radio Resource Management (RRM) and Mobility strategy in E-UTRAN to be based on user specific information by the use of SPID Subscriber Profile ID for RAT/Frequency Priority. It enables individual control of mobility characteristics for a UE based on SPID. The use of SPID makes it possible to create and sell differentiated subscription types. For example the operator can define gold subscribers who have better chance to camp on LTE network (when other RATs are also available) and silver or bronze subscribers that will camp on a 2G or 3G cell even if LTE cell was available. At the same time the network can treat incoming roamers differently compared to own subscribers, or keep M2M subscribers on a specific RAT. A configuration of dedicated frequency at release, target redirect frequency at release (with redirect), target frequency at CS Fallback and target frequency at emergency CS Fallback can be associated with each SPID value. The configuration is done in the enb via the OSS. It is obvious that one or more of the following features must be activated to make use of the feature Subscriber triggered mobility: WCDMA/ GERAN/ CDMA200 session continuity, Coverage triggered inter frequency handover, CS fallback to GSM and WCDMA Ericsson AB 2012 LZT R1A

307 Automated Neighbor Relations Subscriber Profile ID (SPID): index per subscriber, defined in HSS, referring to user information User information can include: mobility profile service usage profile Gives opportunity to sell differentiated subscriptions Enables an operator to differentiate between UEs regarding: Technology/frequency/cell prioritization in IDLE and CONNECTED mode based on SPID Permission to connect to a reserved for operator use cell via incoming handover Figure 7-26: Subscriber Triggered Mobility 3.5 Description The RAT/Frequency Selection Priority (RFSP) values are defined in the HSS and may be modified by MME. The MME translates the RFSP to a SPID value. In the enb the SPID values are mapped to a specific set of RAT/IRAT priorities. The SPID value is transferred from MME to enb at the S1 AP message INITIAL CONTEXT SETUP REQUEST. The enb stores the SPID together with the UE context and it forwards it to target enb over X2 and S1 in case of handover. HSS Subscribers are assigned a SPID value [1..256] MME Sends SPID to enb at context setup and context modification and forwards SPID transparently at S1 handover Removes or adds a specific SPID value for incoming roamers (based on IMSI series analysis) enb OSS configured information in enb Maps SPID to a specific set of RAT/IRAT priorities Sends the SPID value from source to target at handover (both S1 and X2) enb X2 HSS MME S1 S6a enb Figure 7-27: Realization LZT R1A Ericsson AB

308 LTE L13 Radio Network Functionality Associated with each SPID value is a configuration of dedicated frequency at release, target redirect frequency at release, target frequency at CS Fallback and target frequency at emergency CS Fallback. The configurations are received from OSS and are static in enb. If SPID values are configured but enb does not receive any SPID value at INITIAL CONTEXT SETUP REQUEST or at handover the same configuration will be used as if SPID was not configured. There are 256 SPID values. The range is 1-256, values are operator specific and are reserved by 3GPP. Values are predefined. These values can be used by the operator in order to sell differentiated subscriptions and use a specific SPID or RAT/Frequency Selection Priority (RFSP) value for a certain subscription. Total range 1-256, 1-127: operator specific, : reserved by 3GPP enb IDLE and CONNECTED mode configuration for SPID 256 Configuration parameter Value Meaning E-UTRAN carriers priority UTRAN carriers priority GERAN carriers priority high medium low The selection priorities for IDLE and CONNECTED mode of all E-UTRAN carriers are higher than the priorities for all UTRAN and GERAN carriers The selection priorities for IDLE and CONNECTED mode of all UTRAN carriers are lower than the priorities for all E- UTRAN carriers and higher than the priorities for all GERAN carriers The selection priorities for IDLE and CONNECTED mode of all GERAN carriers are lower than the priorities for all E-UTRAN and UTRAN carriers enb configuration for SPID 255 Same as 256 but UTRAN = highest, GERAN = medium, E-UTRAN = low enb configuration for SPID 254 Same as 256 but GERAN = highest, UTRAN = medium, E-UTRAN = low Figure 7-28: 3GPP predefined SPID values: Ericsson AB 2012 LZT R1A

309 Automated Neighbor Relations SPID - 1 LTE F1 LTE F2 WCDMA Normal user Telephony subscription only GSM LTE not included in subscription priority (7 highest) Figure 7-29: Example SPID mapping in enb 3.6 Idle mode mobility Without this feature, the parameter in enb that sets the IDLE MODE mobility priority is called cellreselectionpriority and can be found on the frequency relations in the MOM. The enb collects all cellreselectionpriority from all frequency relations and sends them out to all the UE in the SIB. This means that all UEs get the same information and priorities. Each UE uses these priorities to decide on which RAT and frequency it should camp on when in IDLE mode. Instead, with this feature; SPID lists are used to choose UEs for which to override priority information broadcasted in system information for a configurable period of time. UEs in CONNECTED mode can be configured to override priorities received in System Information for a period of time when in IDLE mode. The UE s SPID is used to select which UEs to configure. The override information can only be sent when UE goes from CONNECTED to IDLE mode in an IE called idlemodemobilitycontrolinfo. The UE (SPID) adapted priority is configured under a new MO Class called RATFreqPrio that apart from the cell reselection priority part also contains a list of SPID values that should use the priorities. LZT R1A Ericsson AB

310 LTE L13 Radio Network Functionality Information sent to UEs only if feature SpidRATFreqPrio is OPERABLE Different priority information sent to UEs based on SPID values RRCConnectionRelease RRCConnectionRelease SPID=3 cellreselectionpriority EUTRANfreq=100 =6 cellreselectionpriority UTRANfreq=250 =2 t320=min20 cellreselectionpriority EUTRANfreq=100 =7 cellreselectionpriority UTRANfreq=250 =5 t320=min30 SPID=12 UEs will use this in IDLEmode Figure 7-30: Idle Mode Mobility 1. Before the RRC connection is released, the enb determines if SPID based priorities should be used for this UE by checking if the SPID value associated with the UE can be found in any of the RATFreqPrio MO. If this is the case, and the feature SpidRATFreqPrio is OPERABLE, then the enb includes the IdleModeMobilityControlInfo IE in the RRC CONNECTION RELEASE message. So, the UE will perform a new cell reselection based on the SPID based priorities (if received). 3.7 Connected Mode Mobility Without this feature, the parameter in enb that sets the CONNECTED MODE mobility priority is called connectedmodemobilityprio and can be found on the frequency/band relations in the MOM. This parameter is used when selecting which frequencies/bands the UE shall perform measurements on. When performing release with redirect and when performing handover, the connectedmodemobilityprio is used to determine the target. When performing CS fallback and CS emergency fallback the enb uses the priority parameter called csfallbackprio and csfallbackprioec to decide which target should be used. All UEs are using the same priority (i.e. will have the same behavior). With this feature, the SPID lists are used to specify groups of UEs to override the priority configuration in the frequency relation MOs. The override priority will affect measurements, release with redirect, handover and CS fallback Ericsson AB 2012 LZT R1A

311 Automated Neighbor Relations Differentiation of UEs done only if feature SpidRATFreqPrio is OPERABLE Configure target measurements Measurements on different frequencies/ technologies configured in the UEs present in different SPID lists SPID=12 Perform meas on EUTRAN freq=100 Perform meas on UTRA freq=150 Perform meas on EUTRAN freq=200 Perform meas on UTRA freq=350 SPID=3 UEs in CONNECTED mode Handover/release/CS Fallback SPID specific target frequency/ technology chosen for a UE only if relation to frequency/cell exists Figure 7-31: Connected Mode Mobility When the feature SpidRATFreqPrio is OPERABLE, the enb configures measurements the UE shall perform based on the connectedmodemobilityprio SPID based, if found in the RATFreqPrio MO. In case of release with redirect or handover, the UE sends a measurement report (bad coverage or best cell). The enb checks whether the UE has a SPID value present in RATFreqPrio. The highest priority that also is a neighbor cell for the current cell is selected as target. In release with redirect the target will be present in the IE RedirectedCarrierInfo. For handover the target will follow the normal handover preparation/execution (but the selection of target is, of course, affected). 3.8 Cell reserved for operator use The feature also includes support for handover restriction in case of cell reserved for operator use. Without the Subscriber Triggered Mobility feature, when a cell is configured in the Cell reserved for operator use state, only UEs that connect to the cell from IDLE mode are denied access. UEs in handover are allowed (this might be needed for testing purposes). The presence of normal customers in the reserved cells can cause problems. With the feature, it is possible to block or admit handover UEs based on their SPID values. The operator can create Handover White lists. The restriction is only valid when the feature is enabled and the cell is configured in the Cell reserved for operator use state. LZT R1A Ericsson AB

312 LTE L13 Radio Network Functionality eg. HO allowed only to SPID 3. HO not allowed SPID 256 SPID 3 HO allowed Cell Reserved for operator use Figure 7-32: Cell Reserved for Operator Use - Incoming Handover 4 Service Triggered Mobility The Service Triggered Mobility feature enables coverage-triggered mobility based on the Quality of Service defined for the UE bearers. The feature applies dynamic levels of coverage thresholds based on the QoS Class Identifier (QCI) profiles of the bearers. With this feature there are different thresholds per QCI for the following events: A2 (bad coverage); A1 (serving cell better than threshold); A5 (target LTE IF cell better than threshold) and B2 (target IRAT cell better than threshold). This is done so that certain services are given higher protection against bad coverage, e.g. to ensure good VoIP quality QCI=1 may have a higher threshold than best effort traffic. One of the following features must be active so that this feature is meaningful: Coverage triggered WCDMA/CDMA2000/ GERAN/ Inter frequency session continuity. Without this feature, the values of the thresholds for A2, A1, A5 and B2 are common for all QCIs. With the feature it is possible to assign specific thresholds for specific QCIs, e.g. if the user has a bearer with QCI=1 it is possible to give this bearer a higher threshold and for example let the bad coverage detection trigger earlier. If a UE is configured with more than one bearer it is the highest threshold that is used to trigger the respective events. The feature is valid for all mobility features using the bad coverage trigger (session continuity and handovers) Ericsson AB 2012 LZT R1A

313 Automated Neighbor Relations Coverage-triggered mobility (session continuity and handovers) is based on the QoS of the UE bearers (QCI). Different thresholds for events A1, A2, A5 and B2 to each UE, based on the UE s bearers QCIs Service based mobility Reconfiguration of UE measurements when: bearer(s) with new QoS profile (QCI) are set up/modified, or last bearer with QCI leading to max thresholds is released Figure 7-33: Service Triggered Mobility There are offset parameters corresponding to all base measurement thresholds for the different measurement events (A1, A2, A5, B2). License control parameters in MO class ServiceTriggeredMobility: featurestateservicetriggeredmobility (Default value 0). There are offset parameters corresponding to all base measurement thresholds for the different measurement events (A1, A2, A5, B2). For each measurement type configured in UE, the measurement thresholds sent to the UE, are computed as the maximum of all threshold values corresponding to the QCI profiles of the UE bearers. Example: a1thresholdrsrpprimoffset is a QCI dependent offset for the absolute threshold a1thresholdrsrpprim. If the value of the offset is 0, then the A1 RSRP threshold corresponding to this QCI profile will be the same as the "base" value in the MO class, namely ReportConfigA1Prim.a1ThresholdRsrpPrim. The rule is that the threshold for a certain QCI profile is computed as: ReportConfigA1Prim.a1ThresholdRsrpPrim + QciProfilePredefined.measReportConfigParams.a1ThresholdRsrpPrimOffset General: eventthresholdxoffset is a QCI dependent offset for the absolute threshold eventthresholdx. If the value of the offset is 0, then the threshold corresponding to this QCI profile will be the same as the "base" value in the MO class, namely ReportConfigEVENT.eventThresholdX. The rule is that the threshold for a certain QCI profile is computed as: ReportConfigEVENT.eventThresholdX + QciProfilePredefined.measReportConfigParams.eventThresholdXOffset LZT R1A Ericsson AB

314 LTE L13 Radio Network Functionality a2thresholdrsrpprim = -138 a5threshold1rsrp = -140 a5threhsold2rsrp = -130 QCI=2 a2thresholdrsrpprimoffset = 5 a5threshold1rsrpoffset = 2 a5threhsold2rsrpoffset = 10 QCI=5 a2thresholdrsrpprimoffset = 3 a5threshold1rsrpoffset = 6 a5threhsold2rsrpoffset = 8 a2thresholdrsrpprimue1=max(a2threhsoldrsrpprim QCI=2,a2ThrehsoldRsrpPrim QCI=5 )= -133 Bearers: QCI=5 QCI=2 a2thresholdrsrpprim QCI=2 = a2threhsoldrsrpprim+ a2thresholdrsrpprimoffset(qci=2)=-133 a2thresholdrsrpprim QCI=5 = a2threhsoldrsrpprim+ a2thresholdrsrpprimoffset(qci=5)=-135 QCI=4 QCI=4 a2thresholdrsrpprimoffset = 1 a5threshold1rsrpoffset = 4 a5threhsold2rsrpoffset = 7 a2thresholdrsrpprimue2 = = -137 Figure 7-34: QCI dependent threshold Settings Whenever one or several bearers are set up, released, or modified for a certain UE for which measurements were previously configured, the feature is used to recompute measurement thresholds due to the bearer QCI constellation change and possibly reconfigure measurements in the UE if some (maximum) measurement threshold value was changed. 5 Redirect with System Information The redirect with system information feature improves outage time when the UE is redirected from LTE a. to GERAN during coverage triggered session continuity, or b. to GSM or WCDMA during CS fallback triggered RRC Connection Release with redirect. (Notice that support for WCDMA comes in L12B when RIM for UTRAN is introduced.) This feature makes the connection to the new RAT much faster as the UE does not have to read system information before accessing the cell. It saves approximately 0.3 sec during Release with Redirect (RwR) to WCDMA and 2sec during RwR to GERAN. Of course the UE must support the e-redirectionutrar9 or e-redirectiongeran-r9 capability (see 3GPP ) Ericsson AB 2012 LZT R1A

315 Automated Neighbor Relations Starting 3GPP release 9 Changes to the Release with Redirect RRC message: Possible to add a list of cells with associated System Information Compatible UE:s will not have to read System Information from target RAT prior to accessing the target cell. Saves approximately 0.3s (UTRAN) / 2s (GERAN) Fully backwards compatible. Figure 7-35: Redirect with system information (NACC) Release with redirect including cellinfolist-r9 (also known as NACC information, or System Information tunneling) to GSM and WCDMA is a mobility procedure from LTE in connected mode to GSM or WCDMA. The total procedure includes bad coverage detection, optionally measurement on target RAT and the release with redirect including the cellinfolist-r9 information element to the UE. The cellinfolist-r9 information element contains system information for a number of potential target cells. An IRAT Release with Redirect with NACC is either blind or based on measurements. Certain configurations in the target network will result in large System Information Containers (SICs), so only a reduced number of cells / SICs will be sent to the UE. We use the coverage indicator to prioritize target cells flagged as "covering". The system information is retrieved by the (RAN Information methods) RIM procedure. Packet core and target RAT must support the RIM procedure for system information. The parameters ExternalGeranCell.rimCapable and ExternalUtranCellFDD.rimCapable indicate if the external GERAN or UTRAN cell is RIM capable. If the value is RIM_Incapable, no attempt will be made to create a RIM association to the external UTRAN cell; other values will result in an attempt to create a RIM association. RIM associations are created when the following criteria are met: GeranCell.rimCapable / ExternalUtranCellFDD.rimCapable is set to RIM_CAPABLE Cell relation exists between a local LTE cell and the external IRAT cell if the cell relation is created by ANR the MO attribute AnrFunctionUtran.rimIntegrationEnabled / AnrFunctionGeran.rimIntegrationEnabled must be set to true. There is no current RIM association between the local cell and target IRAT cell. External IRAT cell is part of a RIM priority pool LZT R1A Ericsson AB

316 LTE L13 Radio Network Functionality Core Network 2 RIM Request UTRA SI / GSM NACC 3 UTRA SI / GSM NACC 5 Attach RBS / BTS* 1 UE attached 6 4 Release with Redirect with System Info Attribute MO class Values enb rimcapable ExternalUtran CellFDD ExternalGeran Cell RIM_CAPABLE *, RIM_INCAPABLE Figure 7-36: Network Assisted Cell Change Only target cells belonging to the redirected frequency will populate the cellinfolist-r9 IE in the RRC Connection Release message. Target cells are selected from available IRAT neighbors in the order shown in the following figure. To reduce enb and core network load two pools per local LTE cell and RAT with potential target cells are created. RIM Session Continuity cell pool Prio 1: target IRAT cell coverage indicator set to COVERS Prio 2: target IRAT cell coverage indicator set to NONE Prio 3: target IRAT cell coverage indicator set to OVERLAPS Prio 4: target IRAT cell coverage indicator set to CONTAINED_IN RIM CS Fallback cell pool Prio 1: target IRAT cell coverage indicator set to COVERS Prio 2: target IRAT cell coverage indicator set to OVERLAPS Prio 3: target IRAT cell coverage indicator set to CONTAINED_IN or NONE * a cell can be present in both pools. Figure 7-37: RIM Priority Pools(included in L12B) The number of cells included in the cellinfolist-r9 IE depends on the size of the system information in the target cells. Up to 4 cells may be included Ericsson AB 2012 LZT R1A

317 Automated Neighbor Relations 6 UE Level Oscillating Handover Minimization Oscillating intra and inter frequency handovers may occur for individual users due to specific radio conditions. Instead of changing cell offset for the cell and impacting all users, this feature will change cell individual offset for each user experiencing oscillating handovers. Non-oscillating UEs are not affected. By reducing the number of unnecessary handovers, the signaling load and potentially the throughput will increase in the network. Oscillating handovers may occur due to specific radio conditions Increased network load (signaling) Increased risk for HO failure Packet loss for UM connections - Voice Reduced throughput Remedy: increased HO margin only for these UEs!!! Figure 7-38: UE Level Oscillating Handover Minimization Based on UE history, information passed on from source enb, the target enodeb will be able to determine if the UE has performed handover to the same cell repeatedly within a short time. When an UE is found to oscillate, the feature will change the handover margin for the UE. This will be enforced when the UE reports mobility measurement and thereby the handover is delayed or prevented. This will decrease the number of unnecessary handovers and reduce the risk for a second handover back to the original source cell. For every UE, the RBS measures the time in current cell (TC), and keeps a record of the time in last cell (TL). This information is used to detect and prevent handover oscillation. The time in current cell (TC) is measured from the time when the Radio Link Control (RLC)/Media Access Control (MAC) layers of the Uu link is established between the UE and the RBS. The time is measured to the handover attempt decision by the RBS, when it has received the Radio Resource Control (RRC) measurement report. The time in last cell (TL) is fetched from the UE History information element provided in the X2 Application Protocol (X2-AP) or S1 Application Protocol (S1-AP) HANDOVER REQUEST message, from the last cell. LZT R1A Ericsson AB

318 LTE L13 Radio Network Functionality At every handover out, the time in current cell TC is measured and compared to limit T2. If TC is less than T2, a handover oscillation has occurred and is counted by the pmhooscinterf or pmhooscintraf Performance Management (PM) counters. The oscillating handover minimization process is triggered by an Event A3 measurement, for intra-frequency relations, or an Event A5 measurement, for inter-frequency relations. The RBS then performs a number of actions. First the TC is calculated, and then the handover oscillation rules are checked. The outcome is either to attempt handover, to reject handover or to apply an offset to the handover margin (HOM), for intra-frequency relations, or Handover Threshold (HOT), for inter-frequency relations. If the decision is to apply an offset the HOM or HOT, the HOM or HOT is calculated the offset is added, and the handover condition is checked. This results in either a decision to attempt handover or to reject handover. Rule 1: Very fast HOs are rejected Probably due to erroneous report from UE Rule 2: Fast HOs are moderated An extra 1 db of gain of HO is required for a fast HO Rule 3: Oscillating HOs are moderated An extra 2 db of gain is required to HO again into a cell the UE didn t stay in last time Figure 7-39: Handover Oscillation Prevention Rules TC TC < T1 T1 TC < T2 T1 TC < T2 T2 TC < T3 T2 TC < T3 T3 TC Time in Last Cell (TL) - TL < T2 TL T2 TL < T2 TL T2 - Action Reject handover. Apply an extra offset2 (2dB) on handover margin for handover back to the last cell. Apply an extra offset1 (1dB) on handover margin for handover to any other cell Apply an extra offset1 (1dB) on handover margin. Apply an extra offset2 (2dB) on handover margin for handover back to the last cell. Attempt handover Attempt handover Attempt handover TC: Time in current cell TL: Time in last cell T1: Time limit 1 = 200 ms T2: Time limit 2 = 2,000 ms T3: Time limit 3 = 5,000 ms Figure 7-40: Handover Oscillation Prevention Rules- details Ericsson AB 2012 LZT R1A

319 Automated Neighbor Relations 7 CS Fallback CS Fallback function in EPS enables the provisioning of the CS services when the UE is served by E-UTRAN. A CS Fallback enabled terminal connected to E- UTRAN can use GERAN or UTRAN to connect to the CS domain. There is also CS Fallback to CDMA2000. Figure 7-41 illustrated reference architecture for the CS Fallback function. EPC CS Core GWMSC GPRS Packet Core GGSN SGSN S4 S6d S3 S6a HSS P-GW MSC SGs MME S11 S5/S8 S-GW S1-CP S1-UP GERAN UTRAN E-UTRAN X2-UP Uu enodeb X2-CP enodeb Figure 7-41 CS Fallback to GERAN/UTRAN The CS Fallback and SMS over SGs in EPS are realized by using SGs interface between Control Plane nodes MME and MSC Server. The main purpose with SGs interface is to provide mechanisms for mobility management and paging procedures between EPS domain and CS Domain. It is also used for signaling needed for both MO and MT procedures. LZT R1A Ericsson AB

320 LTE L13 Radio Network Functionality In addition there is a reference model for CS Fallback to CDMA2000 network as illustrated in Figure EPC HSS S6a P-GW S2a PDSN CDMA2000 HRPD (EV-DO) 1XCS MSC MME S11 S5/S8 S101 S103 RNC 1XCS IWS S-GW S1-CP S1-UP A1 E-UTRAN CDMA2000 Access X2-UP Uu enodeb X2-CP enodeb Tunneled 1xRTT message Figure 7-42 CS Fallback to 1xRTT CDMA2000 The CS Fallback for 1xRTT in EPS enables the delivery of CS domain services by reuse of 1xCS infrastructure when the UE is served by E-UTRAN. The CS fallback to 1xRTT and SMS over S102 in EPS function is realized by reusing the S102 reference point between the MME and the 1xCS IWS. The S102 reference point provides a tunnel between MME and 3GPP2 1xCS IWS to relay 3GPP2 1xCS signaling messages. 7.1 Why CS Fallback? LTE/SAE provides connectivity towards PS domain and supports packet based services only. That implies that traditional CS services such as CS Voice, CS Data, SMS are not supported. In order to provide smooth migration CS Fallback solution is provided by 3GPP. A CS fallback enabled terminal, connected to E UTRAN may use GERAN/UTRAN or CDMA 1xRTT to connect to the CS domain. This function is only available in case E UTRAN coverage is overlapped by either GERAN/UTRAN or CDMA coverage. CS Fallback may be used as a generic telephony fallback method securing functionality for incoming roamers as well Ericsson AB 2012 LZT R1A

321 Automated Neighbor Relations Subscribers roaming with preference on LTE access, no CS-voice service available (i.e. IMS is not used as voice engine) Fallback triggered to overlapping CS domain (2G/3G) whenever voice service is requested Resumed LTE access for PS services after call completion (cell reselection) LTE island PS PS LTE CS (+PS) LTE LTE GERAN/UTRAN LTE Figure 7-43 CS Fallback Concept As illustrated in Figure 7-43 LTE coverage will initially only be deployed in islands. Outside these islands, the subscriber must receive its services from a non LTE environment. This can either be a HSPA network, over which MMTel is run or a classical CS network without MMTel capabilities. 2. CS domain updated of subscribers whereabouts through CS signaling over MME-MSC (LUP, SMS etc.) 1. Subscriber registered in MSC but roam in LTE CS signaling Packet Core LTE RAN CSFB Terminal MME SGSN SAE Gw GGSN 4. Page over SGs-interface MSS GSM / WCDMA RAN 5. RAN triggers an release with redirect CSFB Terminal 6. Page response and call setup over 2G/3G radio payload RC M-MGw IM-MGw MRFP MSC-S MGCF 3. Incoming call to subscriber in LTE Figure 7-44 MSS as voice engine for LTE subscribers Figure 7-44 illustrates one scenario when providing CS services to an LTE subscriber. Prerequisite for CS Fallback to work is that UE is dual radio capable UE and it is registered in the CS Domain. LZT R1A Ericsson AB

322 LTE L13 Radio Network Functionality 7.2 Emergency Call for CS Fallback WCDMA network as CS fallback for ordinary calls GSM network is preferred for emergency calls Frequency Relation: CS FallbackPriority Emergency Call [7-0] CS 112 UTRAN GRAN Figure 7-45 Emergency Call Support for CS Fallback The Emergency Call Support for CS Fallback feature enables the operator to set unique frequency and frequency group priorities for emergency calls. The operator might prefer to use the WCDMA network as CS fallback for ordinary calls while the GSM network is preferred for emergency calls Ericsson AB 2012 LZT R1A

323 Automated Neighbor Relations 8 Neighbor Cell Relations In an LTE network, neighbor cell relationships can be constructed: Manually, loading neighbor cell relations using OSS-RC. Automatically, building a cell neighbor list using ANR, see chapter 7. The figure below illustrates managed object structure. EUtranCellFDD Up to 32 Cdma2000FreqBandRelation Up to16 UtranFreqRelation Up to16 GeranFreqGroupRelation Up to 8 EUtranFreqRelation Up to 64 CDMA2000CellRelation UtranCellRelation Up to 64 Up to 64 GeranCellRelation Up to 64 EUtranCellRelation External Internal cellindividualoffset AnrCreated... Figure 7-46: Neighbor Cell Relations MOM view Neighbor Cell Relations are defined using following MOs: EUtranFreqRelation for Intra LTE Neighbors GERANFreqGroupRelation for GSM neighbors UTRANFreqRelation for WCDMA neighbors CDMA2000FreqBandRelation for CDMA2000 neighbors LZT R1A Ericsson AB

324 LTE L13 Radio Network Functionality The system information blocks SIB5, SIB6, SIB7 and SIB8 conveys the information necessary for the UE to perform IRAT and Inter Frequency mobility functions. SIB5 - EUTRA The redirection priorities set per cell: connectedmodemobilityprio The priorities can be set for : GeranFreqGroup UtranFrequency Cdma2000Freq EUtranFreqency enodeb IRAT/IF SIB SIB6 - UTRA SIB5 threshxhigh SIB7 - GRAN threshxlow SIB8 CDMA2000 SIB8 allowedmeasbandwidth treselectioncdmahrpd qoffsetfreq treselectioncdmahrpdsfhigh SIB7 presenceantennaport1 treselectioncdmahrpdsfmedium treselectiongran SIB6 hrpdcellreselectionpriority treselectiongransfhigh treselectionutra hrpdbandclass treselectiongransfmedium treselectionutrasfhigh pmaxgeran cellreselectionpriority treselectionutrasfmedium treshxhighhrpd qrxlevmin treshxlowhrpd pmaxgeran freqcdma treshx-high pnoffset treshx-low csfbsupportforlimiteddualradioues Figure 7-47: IRAT/IF Mobility: Measurement Configuration (GSM, WCDMA, CDMA2000 and LTE Inter Frequency) 9 Inter-frequency Load Balancing Load is balanced between overlapping cells. The load per cell is calculated and exchanged between cells with overlapping coverage. If load imbalance exists a number of UEs will be moved to lower loaded LTE carrier based on handover. This feature was first introduced in L12B and is dependant on: Coverage-Triggered Inter-Frequency Handover 9.1 Summary The feature distributes traffic load between inter-frequency (LTE) cells that cover the same geographical area. Users in connected mode are distributed by loadtriggered inter-frequency handovers. Consistent behavior in idle mode is ensured by sticky carrier. Important to highlight: Arbitrary grade of coverage overlap between cells is supported Ericsson AB 2012 LZT R1A

325 Automated Neighbor Relations An arbitrary number of cells on the same target inter-frequency is supported. An arbitrary number of frequencies (and frequency bands) is supported. Load balancing between co-located and non-co-located cells are supported. Load balancing between intra-enb and inter-enb cells are supported. 9.2 Description Cells with coverage overlap are candidates for inter-frequency load balancing. Cells with coverage overlap are configured as cells with load relation. Idle mode priorities broadcasted in system information is arranged to have the highest priority, among frequencies part of the load relation, for the camped frequency. As a consequence a UE camping on one frequency will be locked in to this frequency, independent which frequency camping on. Connected to idle transitions will not trigger the UE to re-select cell. Cells with load relation exchange load information. Higher loaded cells will select a number of UEs as candidates for load balancing. Selected UEs are configured for measurement on lower loaded cell and a number of the UEs will report to be in coverage of the lower loaded cell. A number of UEs, corresponding to the imbalance, will be moved to the lower loaded cell by PS handover. The whole cycle is repeated a few times per minute. The feature is supported between cells within an Ericsson enb and between cells in separate Ericsson enbs. The feature distributes traffic load between inter-frequency (LTE) cells that cover the same geographical area. Users in connected mode are distributed by load-triggered interfrequency handovers. Consistent behavior in idle mode is ensured by sticky carrier. Frequency Distance Figure 7-48: Inter Frequency Load Balancing - Feature Description LZT R1A Ericsson AB

326 LTE L13 Radio Network Functionality The main steps on how the feature works is illustrated in the figure below. Load measurement Determine traffic load in each individual cell Load information exchange Request and exchange load information between cells Load balancing action magnitude Determine the amount of off-loading needed to balance load between cells UE selection for load balancing action Determine and off-load suitable UEs Figure 7-49: Functional Overview 9.3 Feature Benefits: Capacity: The feature distributes UEs so that the total traffic capacity is the aggregated sum of traffic capacity for each individual inter-frequency cell covering the same area. Examples of traffic related capacities are system throughput and number of UEs and bearers. Performance: The feature increases the average user throughput performance linearly with increased bandwidth when bandwidth is increased by adding cell carriers and when load (total number of users) is constant. Fairness: The feature enables users to experience similar performance independent of used cell carrier Ericsson AB 2012 LZT R1A