R&S TSMW, TSME, TSMA LTE Downlink Allocation Analysis Application Note

Transcription

1 R&S TSMW, TSME, TSMA LTE Downlink Allocation Analysis Application Note Products: ı ı ı ı R&S TSMW R&S TSME R&S TSMA R&S ROMES4 Application Note Jordan Schilbach

2 Table of Contents 1Introduction 3 LTE Downlink Allocation Analysis 1.1Network Optimization - Throughput Testing Network Optimization - Site Assessment Network Benchmarking Achieving Results with DLAA Cell Level statistical results UE level statistical results Complete list of statistical results Carrier Aggregation and Cross carrier scheduling Carrier Aggregation Cross-carrier scheduling Supported Allocations Ordering Information Rohde & Schwarz LTE Downlink Allocation Analysis 2

3 1 Introduction The R&S TSMW, TSME or TSMA in combination with the K31 software options offer a unique feature that allows the analysis of the downlink allocations of the LTE enodebs during measurement. This information is normally only available from the inside of the mobile network, provided by the OSS. Now this information can be made available during a drive test on the go, to allow a deep investigation of the network performance. The following are the main applications. 1.1 Network Optimization - Throughput Testing Typically throughput testing is used to verify the network performance. If the UE throughput is high the optimization engineer can be sure that the network is in good shape. However if the throughput is low, this can be caused by many issues, a simple reason could be that the network is heavily loaded. This information is not available to the engineer in the field, or often not even in the backoffice. The DLAA provides this information directly in the drive test tool, so the engineer can see that a low testmobile throughput is caused by network congestion, and does not point to a failure in the network itself. 1.2 Network Optimization - Site Assessment Network Planning and optimization strives to maximize spectral efficiency and user experience by providing capacity where it is actually needed. ı Cells that have to serve too many users can only provide a degraded quality of service. ı At the same time a cell's performance will suffer if there are too many users at the cell edge. Both effects can be identified by the DLAA results. The DLAA can assess the number of RNTIsnumber of UEs in a cell, as well as the Modulation and Coding Scheme (MCS) that is used. This helps the network planner to identify overloaded cells, as well as inefficiently used cell. So the DLAA can be used as a tool to plan investments in cell capacity, such as adding carriers, cell splitting or adding small cells.

4 1.3 Network Benchmarking In addition to traditional benchmarking the DLAA feature provides additional insights into the networks. Complementary to the throughput tests, the DLAA can now measure the load of the cells, and estimate the number of subscribers in a cell. This allows the benchmarking application to provide a detailed comparison between operators regarding subscriber numbers, average subscriber throughput and cell throughput.

5 2 Achieving Results with DLAA The Downlink Allocation Analysis is performed in parallel to the Wideband measurements. Therefore it is important to activate wideband measurements. LTE Resource Allocation Scheme: Source:

6 The DLAA analyses the enodeb allocations to the UEs in the cell for a certain analysis time. In the above example, the analysis is done for a measurement time of 10 subframes (10 ms). Only 3 MHz of the carrier bandwidth are shown. The minimum amount of data that can be scheduled to a UE is 1 TTI (1ms) and 1 RB (blue square). The DLAA analyses all TTIs and RBs in the measurement time, and provides detailed allocation information such as: RNTI Modulation and Coding Scheme DCI format Transport Blocksize System Frame Number Subframe Number The DLAA analysis is done with the same rate as the wideband scanning is performed, hence the number of DLAA results can be controlled by adjusting the WB measurement rate. The measurement time is automatically chosen by the system to be as short as possible, to achieve a high measurement rate. However the user can increase the measurement time in the advanced settings of the WB scanner (in this example 50ms): The measurement time is visible in the LTE Scanner DLAA view, as the number of TTIs between the black vertical lines: In this example there are 22 TTIs between the black vertical lines, i.e. the measurement time was 22ms. Black lines are visible in this view with the following configuration settings:

7 The measurement rate is the WB measurement rate: The measurement rate indicates how many results, each of a duration of the measurement time described above, have been measured per second. Based on the raw results from the DLAA, ROMES and the NPA calculate measurement results by interpolating and averaging the allocations over time. There are cell level and UE level results available. 2.1 Cell Level statistical results Cell level results are available per cell, and hence can be viewed in the LTE Scanner TopN view, and the LTE Scanner signal tree. Several averaging windows (known as observation intervals) with different lengths can be defined under technology for LTE scanner: The main results calculated from the raw results are 1. Cell throughput Based on the raw data, the system knows the exact allocated amount of data for the measurement time. From this data, the cell throughput can be calculated. A minimum, maximum and average throughput is provided. 2. RB usage in %

8 Based on the raw data, the system knows the exact number of RBs that were used for data transmission and hence the RB usage in percent. From this data, the RB utilization can be calculated. 3. number of RNTIs (correlates with number of UEs) An RNTIs is assigned to the UE, when data needs to be transmitted. When the transmission ends, the RNTI is released after the expiry of a timer (typically 15 seconds). A UE may have several RNTIs at the same time. Hence the number of RNTIs detected in a cell strongly correlates with the number of users in the cell. The interpolation algorithm will count the different detected RNTIs within the observation interval. As the observation interval can be defined by the user, the algorithm can be calibrated to achieve a good match to the RNTI connected user counter of the enodeb. 4. RNTI throughput The RNTI throughput is calculated from the cell throughput divided by the average of RNTIs, and correlates strongly with the UE throughput in the cell. A minimum, maximum and average throughput is provided. The signal is called DLAA Avg. Sched. TP (UE). 5. Subframe usage in % This is the number of all subframes which contain at least one resource block witch scheduled data divided by the number of all measured subframes 6. Average MCS This is the average MCS over all allocations in the cell. Comparing the average MCS between different cells or operators allows to assess if the cell is at the right location for the offered user traffic, and hence if spectrum resources are used effectively. To analyze these results, they need to be activated in the Technology for LTE scanner TopN signal settings: The results are available in the signal tree:

9 2.2 UE level statistical results UE level results represent information for a single RNTI in the DLAA RNTI statistics view. The values are averaged over a user defined observation interval. The interval and the number of UEs to display in the view can be configured in the view's configuration dialog: The view presents the result in a barchart for each RNTI. The view shows the RNTIs with the highest result values, for the value selected by the user by clicking on the value in the legend. In this example, the view shows 4 UEs, and the view is sorted by scheduled throughput.

10 The following results are available on UE level: 1. RB Usage [%] Sum of all used RBs of the UE in the measured subframes sum of all RBs in the measured subframes is an indicator how much resource the UE consumes on average in the observation interval 2. RB Usage when scheduled [%] Sum of all used RBs of the UE in the measured Sub Frames Sum of all RBs in the measured Subframes which contain at least one RB for the UE, is an indicator how much resources the UE consumes when it is actually actively receiving data 3. Average MCS Average MCS of the RNTI. Indicator if the specific UE is effectively using the cell resources. UEs at cell edge will have lower MCS, and use resources less effectively. 4. Scheduled throughput Amount of data scheduled for the UE divided by the measurement time. Is an indicator for the average throughput the UE receives from the network. 5. Scheduled throughput when scheduled Amount of data scheduled for the UE divided by the duration of those subframes, where the UE actually recieves data from the enodeb. Is an indicator for the peak throughput the UE receives from the network when actively being scheduled. 6. Subframe Usage Sum of all Sub Frames which contain at least one RB for the UE Sum of all Measured Sub Frames. Is an indicator for the activity of the UE. 7. Average Number of Transport Blocks When MIMO is in use, the UE will receive 2 transport block in a subframe. Hence the average count of transport blocks of a UE is an indicator for MIMO usage: 1 is for SISO, 2 is for MIMO, and values in between indicate that MIMO is used only on a subset of RBs. 2.3 Complete list of statistical results Parameter Calculation Explanation ViewSignalusage Number of RNTIs Number of different RNTIs found in the DLAA results during the Observation interval Signal DLAA Num. RNTI Number of RNTIs

11 The number of RNTIs found by the DLAA for a ChannelPCI RB usage of UE RB usage when Scheduled of UE RB usage of Cell Sched. TP of UE Sched. TP of UE when Scheduled Average Scheduled throughput of UE [Sum of all used RBs of the UE in Measured Sub Frames] [Sum of all RBs in Measured Subframes], [Sum of all used RBs of the UE in Measured Sub Frames] [Sum of all RB in Measured Subframes which contain at least one RB for the UE] [Sum of all allocated RBs in Measured Sub Frames] [Sum of all RB in Measured Subframes] Sum of all Transport Block Sizes of the UE (n * (0.001s)) in [bitss] Where n is the number all measured sub frames in the DLAA observation period. [Sum of all Transport Block Sizes of the UE] [(m * (0.001s))] in [bitss] Where m is the number of all measured sub frames which contain at least one RB for the UE in the DLAA observation period. DLAA Avg. Sched. TP (Cell) DLAA Num. RNTI RNTI statistics view Indicator how much resources the UE consumes on average in the observation interval RNTI statistics view Indicator how much resources the UE consumes when it is actually actively receiving data Signal DLAA RB Usage Resource Block Usage Percentage of the resource blocks which are allocated to all UEs RNTI statistics view Amount of data scheduled for the UE divided by the measurement time. Is an indicator for the average throughput the UE receives from the network. RNTI statistics view Amount of data scheduled for the UE divided by the duration of those subframes, where the UE actually recieves data from the enodeb. Is an indicator for the peak throughput the UE receives from the network when actively being scheduled. Signal DLAA Avg. Sched. TP (UE)

12 Average. Sched. TP of Cell [Sum of all Transport Block Sizes] [(n * (0.001s))] in [bitss] Where n is the number all measured sub frames in the DLAA observation period Avg. Scheduled Throughput (UE) Average Throughput of the UEs in an enodeb, calculated by the average throughput of the enodeb diveded by the number of UEs in the enodeb in the observation interval Note: in addition to the average value, also minimum and maximum values are available. Signal DLAA Avg. Sched. TP (Cell) Avg. Scheduled Throughput Average scheduled Throughput of the enodeb, calculated by the usage of resource blocks at the TTIs and the reported transport block size. Note: in addition to the average value, also minimum and maximum values are available. Average MCS of UE Average MCS of Cell [Sum of all MCS of the UE] [Number of MCS] in the DLAA observation period. [Sum of all MCS] [Number of MCS] in the DLAA observation period RNTI statistics view Indicator if the specific UE is effectively using the cell resources. UEs at cell edge will have lower MCS, and use resources less effectively. Signal DLAA Avg. MCS Average MCS of the enodeb. This is the sum of all MCS divided by the number of MCS in the observation period. 8. Comparing the average MCS between different cells or operators allows to assess if the cell is at the right location for the offered user traffic, and hence if spectrum

13 resources are used effectively. Note: in addition to the average value, also minimum and maximum values are available. Sub Frame usage of UE [Sum of all Sub Frames which contain at least one RB for the UE] [Sum of all Measured Sub Frames] RNTI statistics view Is an indicator for the activity of the UE. Sub Frame usage of Cell Average Count of Transport Blocks of UE when Scheduled [Sum of all Sub Frames which contain at least one RB] [Sum of all Measured Sub Frames] [Number of TBs of the UE] [Total number of TTIs] in the DLAA observation period Signal DLAA Subframe Usage (Cell) This is the number of all subframes which contain at least one resource block with scheduled data divided by the number of all measured subframes RNTI statistics view When MIMO is in use, the UE will receive 2 transport block in a subframe. Hence the average count of transport blocks of a UE is an indicator for MIMO usage: 1 is for SISO, 2 is for MIMO, and values in between indicate that MIMO is used only on a subset of RBs.

14 3 Carrier Aggregation and Cross carrier scheduling 3.1 Carrier Aggregation With LTE-A it is possible to schedule data to a UE from multiple carriers using the carrier aggregation feature. When a UE is scheduled on both carriers, the same RNTI is used on both carriers. ROMES detects from the DLAA results, in which set of carriers in an operators network CA is active. The CA detection is displayed in the RNTI Table View. 3.2 Cross-carrier scheduling When cross carrier scheduling is used in the network, the scheduling for the UE on a secondary component carrier (SCC) is signalled on the primary component carrier (PCC). The scanner can detect such allocation scheduling information if PCC and SCC use the same carrier bandwidth, but it can not identify for which carrier such allocations are scheduled. If cross carrier scheduling is detected, the RNTI column in the RNTI Statistics View is marked with a *, e.g 7800* and a remark Cross Carrier Aggregation will be displayed in column CA EARFCNs of the new RNTI Table View.

15 Example: carrier A and carrier B have the same bandwidth and the network uses cross carrier scheduling as well as own carrier scheduling. Cross carrier scheduling allocations for carrier B are signalled on carrier A. The scanner can detect that carrier A includes allocation information for another carrier, but the scanner can not detect that the allocation is for carrier B. In consequence, the throughput, load and RNTI allocation analysis for carrier B will only show results from allocations that are signalled on carrier B, but will not include the allocations decoded on carrier A from cross carrier scheduling. Therefore, the results for carrier B will be lower than what they are in reality.

16 4 Supported Allocations DCI Format Number of Tx Antennas Channel Type 1 1 SIMO Yes 1A 1 or 2 SIMO or MISO Yes Supported Allocation Analyzer in ROMES B1D 2 or 4 MISO 2 and 4 Tx supported 2 2 or 4 MIMO 2 and 4 Tx supported 2A 2 or 4 MIMO 2 and 4 Tx supported 2B 2 MU-MIMO or SU-MIMO No 2C 2 to 8 MU-MIMO or SU-MIMO No 0 1 SISO No 4 2 or 4 MIMO No For information:

17 5 Ordering Information Designation Type Order No. ROMES4 Drive Test SW R&S ROMES ROMES4 Scanner Driver for TSMETSMA R&S ROMES4T1E ROMES4 Scanner Driver for TSMW R&S ROMES4T1W LTE Downlink Allocation Analyzer option for TSMW R&S TSMW-K LTE Downlink Allocation Analyzer option for TSME R&S TSME-K LTE Downlink Allocation Analyzer option for TSMA R&S TSMA-K

18 About Rohde & Schwarz The Rohde & Schwarz electronics group offers innovative solutions in the following business fields: test and measurement, broadcast and media, secure communications, cybersecurity, radiomonitoring and radiolocation. Founded more than 80 years ago, the independent company which is headquartered in Munich, Germany, has an extensive sales and service network with locations in more than70 countries. Mobile network testing The company s broad and diverse product portfolio for mobile network testing addresses every test scenario in the network lifecycle from base station installation to network acceptance and network benchmarking, from optimization and troubleshooting to interference hunting and spectrum analysis, from IP application awareness to QoS and QoE of voice, data, video and app-based services. Regional contact Europe, Africa, Middle East customersupport@rohde-schwarz.com North America TEST RSA ( ) customer.support@rsa.rohde-schwarz.com Latin America customersupport.la@rohde-schwarz.com Asia Pacific customersupport.asia@rohde-schwarz.com China customersupport.china@rohde-schwarz.com Sustainable product design ı Environmental compatibility and eco-footprint ı ı Energy efficiency and low emissions Longevity and optimized total cost of ownership This and the supplied programs may only be used subject to the conditions of use set forth in the download area of the Rohde & Schwarz website. R&S is a registered trademark of Rohde & Schwarz GmbH & Co. KG; Trade names are trademarks of the owners. PAD-T-M: EN Rohde & Schwarz GmbH & Co. KG Mühldorfstraße Munich, Germany Phone Fax