FibeAir IP-10Q. Product Description. ANSI Version. April 2013 Hardware Release: R1 Software Release: Q6.9 Document Revision B.02

Transcription

1 ANSI Version FibeAir IP-10Q April 2013 Hardware Release: R1 Software Release: Q6.9 Document Revision B.02 Copyright 2013 by Ceragon Networks Ltd. All rights reserved.

2 Notice This document contains information that is proprietary to Ceragon Networks Ltd. No part of this publication may be reproduced, modified, or distributed without prior written authorization of Ceragon Networks Ltd. This document is provided as is, without warranty of any kind. Registered Trademarks Ceragon Networks is a registered trademark of Ceragon Networks Ltd. FibeAir is a registered trademark of Ceragon Networks Ltd. CeraView is a registered trademark of Ceragon Networks Ltd. Other names mentioned in this publication are owned by their respective holders. Trademarks CeraMap, PolyView, EncryptAir, ConfigAir, CeraMon, EtherAir, and MicroWave Fiber, are trademarks of Ceragon Networks Ltd. Other names mentioned in this publication are owned by their respective holders. Statement of Conditions The information contained in this document is subject to change without notice. Ceragon Networks Ltd. shall not be liable for errors contained herein or for incidental or consequential damage in connection with the furnishing, performance, or use of this document or equipment supplied with it. Open Source Statement The Product may use open source software, among them O/S software released under the GPL or GPL alike license ("GPL License"). Inasmuch that such software is being used, it is released under the GPL License, accordingly. Some software might have changed. The complete list of the software being used in this product including their respective license and the aforementioned public available changes is accessible on Information to User Any changes or modifications of equipment not expressly approved by the manufacturer could void the user s authority to operate the equipment and the warranty for such equipment. Ceragon Proprietary and Confidential Page 2 of 225

3 Revision History Rev Date Author Description Approved by Date A February 28, 2012 Baruch Gitlin First revision for release 6.9. Tomer Carmeli February 28, 2012 A.01 March 11, 2012 Baruch Gitlin Revised description of encryption algorithms for secure management protocols. Nir Gasko March 11, 2012 A.02 March 15, 2012 Baruch Gitlin Revise PDV value for PTP optimized transport. Tomer Carmeli March 15, 2012 A.03 March 22, 2012 Baruch Gitlin Updated RFU-C frequency specs. Rami Lerner March 26, 2012 A.04 April 2, 2012 Baruch Gitlin Revise RFU-C frequency specifications. A.05 July 4, 2012 Baruch Gitlin Revise RFU environmental specifications. Rami Lerner April 2, 2012 Rami Lerner July 4, 2012 A.06 July 12, 2012 Baruch Gitlin Revise capacity specifications. Eran Shecter July 12, 2012 B September 19, 2012 Baruch Gitlin Added technical details, and revised document structure and format. Eran Shecter September 23, 2012 B.01 October 29, 2012 Baruch Gitlin Updated RFU-C mediation device losses. Eran Shecter/Rami Lerner October 29, 2012 B.02 April 29, 2013 Baruch Gitlin Revise RFU-C Transmit Power specifications for 24 GHz UL. Rami Lerner April 29, 2013 Ceragon Proprietary and Confidential Page 3 of 225

4 Table of Contents 1. Synonyms and Acronyms Introduction Product Overview IP-10Q Advantages Efficient Utilization of Spectrum Assets Spectral Efficiency Radio Link Wireless Network Scalability Availability Network Level Optimization Network Management Power Saving Mode with High Power Radio Functional Description Functional Overview Single Point of Management Centralized System Features Solution Overview Hardware Description Hardware Architecture Front Panel Description Ethernet Interfaces GbE Interfaces Management Interfaces Radio Interface Power Interfaces Additional Interfaces Front Panel LEDs External Alarms Licensing License Overview Working with License Keys Licensed Features Feature Description Equipment Protection Equipment Protection Overview HSB Protection Multi-Radio and 2+0 Multi-Radio with IDU and Line Protection Switchover Triggers Ceragon Proprietary and Confidential Page 4 of 225

5 5.2 Ethernet Line Protection Ethernet Line Protection Options Multi-Unit LAG Capacity and Latency Capacity Summary Ethernet Header Compression Latency Radio Features Adaptive Coding Modulation (ACM) ACM with Adaptive Transmit Power Cross Polarization Interface Canceller (XPIC) Multi-Radio Diversity ATPC Override Timer Ethernet Features Automatic State Propagation Quality of Service (Traffic Manager) Integrated Quality of Service (QoS) Overview Standard QoS Enhanced QoS Standard and Enhanced QoS Comparison Synchronization Synchronization Overview IP-10Q Synchronization Solution Synchronization Using Precision Timing Protocol (PTP) Optimized Transport SyncE PRC Pipe Regenerator Mode Radio Frequency Units (RFUs) RFU Overview RFU Selection Guide RFU-C Main Features of RFU-C RFU-C Frequency Bands RFU-C Mechanical, Electrical, and Environmental Specifications RFU-C Mediation Device Losses RFU-C Antenna Connection RFU-C Waveguide Flanges HP/RFU-HP Main Features of 1500HP/RFU-HP HP/RFU-HP Frequency Bands HP/RFU-HP Mechanical, Electrical, and Environmental Specifications HP/RFU-HP Functional Block Diagram and Concept of Operation HP/RFU-HP Comparison Table HP/RFU-HP System Configurations HP/RFU-HP Space Diversity Support Split Mount Configuration and Branching Network Split-Mount Branching Loss HP/RFU-HP All Indoor Configurations and Branching Network HP/RFU-HP All Indoor Compact (Horizontal) HP/RFU-HP Models and Part Numbers Ceragon Proprietary and Confidential Page 5 of 225

6 OCB Part Numbers Generic All-Indoor Configurations Part Numbers RFH-HS Main Features of RFU-HS RFU-HS Frequency Bands RFU-HS Mechanical, Electrical, and Environmental Specifications RFU-HS Antenna Types RFU-HS Antenna Connection RFU-HS Mediation Device Losses RFU-SP Main Features of RFU-SP RFU-SP Frequency Bands RFU-SP Mechanical, Electrical, and Environmental Specifications RFU-SP Direct Mount Installation RFU-SP Antenna Connection RFU-SP Mediation Device Losses P P Mechanical, Electrical, and Environmental Specifications P Mediation Device Losses Typical Configurations Single Pipe Configurations Supported Configurations Single Pipe Seamless Upgradeability Configuration Multi-Radio Configuration Multiple Pipe Configurations (Chain/Node Sites) Seamless Upgradeability Multiple Pipes Multi-Radio East/West Configuration FibeAir IP-10Q Management Management Overview Management Communication Channels and Protocols Web-Based Element Management System (Web EMS) Command Line Interface (CLI) Text CLI Configuration Scripts Floating IP Address In-Band Management In-Band Management Isolation Out-of-Band Management System Security Features Ceragon s Layered Security Concept Defenses in Management Communication Channels Defenses in User and System Authentication Procedures Secure Communication Channels Security Log Ethernet Statistics Ceragon Proprietary and Confidential Page 6 of 225

7 8.9.1 Ingress Line Receive Statistics Ingress Radio Transmit Statistics Egress Radio Receive Statistics Egress Line Transmit Statistics Radio Ethernet Capacity Radio Ethernet Utilization Software Update Timer CeraBuild Standards and Certifications Carrier Ethernet Functionality Supported Ethernet Standards Standards Compliance Network Management, Diagnostics, Status, and Alarms Specifications General Specifications GHz GHz Transmit Power Specifications RFU-C Transmit Power (dbm) HP/RFU-HP Transmit Power (dbm) RFU-HS Transmit Power (dbm) RFU-SP Transmit Power (dbm) P Transmit Power (dbm) Receiver Threshold Specifications RFU-C Receiver Threshold (RSL) BER = 10-6) HP/RFU-HP Receiver Threshold (RSL) = 10-6) RFU-HS Receiver Threshold (RSL) BER = 10-6) RFU-SP Receiver Threshold (RSL) BER = 10-6) P Receiver Threshold (RSL) BER = 10-6) Radio Capacity Specifications Radio Capacity without Header Compression Radio Capacity with Legacy MAC Header Compression Radio Capacity with Multi-Layer Enhanced Header Compression Ethernet Latency Specifications Latency 30MHz Channel Bandwidth Latency 40MHz Channel Bandwidth Latency 50MHz Channel Bandwidth Latency 56MHz Channel Bandwidth (for 80MHz channels) Interface Specifications Mechanical Specifications Power Input Specifications Power Consumption Specifications Power Consumption with RFU-HP in Power Saving Mode Environmental Specifications Ceragon Proprietary and Confidential Page 7 of 225

8 List of Figures Functional Block Diagram FibeAir IP-10Q Block Diagram IP-10Q Chassis - Module Numbering IP-10Q Front Panel and Interfaces HSB Node with BBS Space Diversity Multi-Radio 2+0 with Line Protection Traffic Flow Hardware Protection with Single Interface Using Optical Splitter Full Protection Using Multi-Unit LAG Multi-Unit LAG Basic Operation Layer 1 Header Suppression Legacy MAC Header Compression Multi-Layer (Enhanced) Header Compression Adaptive Coding and Modulation with Eight Working Points Adaptive Coding and Modulation IP-10Q ACM with Adaptive Power Contrasted to Other ACM Implementations69 Dual Polarization XPIC - Orthogonal Polarizations XPIC Impact of Misalignments and Channel Degradation XPIC Impact of Misalignments and Channel Degradation Typical 2+0 Multi-Radio Link Configuration Direct and Reflected Signals Diversity Signal Flow QoS Traffic Flow IP-10Q Enhanced QoS Classifier Traffic Flow Synchronized Packet Loss Random Packet Loss with Increased Capacity Utilization Using WRED WRED Profile Curve Queue Priority Configuration Example Example 1 Hybrid Scheduling Illustration Ceragon Proprietary and Confidential Page 8 of 225

9 Example 1 Hierarchical Scheduling Illustration Precision Timing Protocol (PTP) Synchronization Synchronous Ethernet (SyncE) Figure 1: 1500HP 2RX in 1+0 SD Configuration Figure 2: 1500HP 1RX in 1+0 SD Configuration Space Diversity with Multiple RFUs Space Diversity with Single RFU All-Indoor Vertical Branching Split-Mount Branching and All-Indoor Compact Old OCB New OCB Old OCB Type Old OCB Type 1 and Type 2 Description Block Diagram of Trunk System All-Indoor System with Five IP-10 Carriers All-Indoor System with Ten IP-10 Carriers All-Indoor Installations Subrack for ETSI Rack RFU with Branching ICB Branching Chain ICC ICCD Fan Tray in 19 Frame Rack T12 Rigid Waveguide T13 Rigid Waveguide XPIC Assembly Configuration Additional Assembly Configuration Examples Lab Rack (Open Frame) Examples Rack Example ETSI Rack Example Configuration with More than Ten Carriers Two Connected Racks Ceragon Proprietary and Confidential Page 9 of 225

10 1500HP RFU All-Indoor 1Rx RF Unit HP RFU All-Indoor Space Diversity HP RFU All-Indoor 1Rx RF Unit, 11G 40MHz HSB Compact Front View HSB Compact Rear View PDU with 10 Switches PN: 32T-PDU Seamless Upgradeability Single Pipe FibeAir IP-10Q Typical Configurations FibeAir IP-10Q Typical Configurations 2+0 Multi-Radio Seamless Upgradeability Multiple Pipes Seamless Upgradeability Multiple Pipes- 2+0 and 2 x Up to 2 x 2+0 Multi-Radio (XPIC Optional) and 1+1 HSB (BBS Space Diversity Optional) Multi-Radio East/West Configuration Integrated IP-10Q Management Tools In-Band Management Isolation Security Solution Architecture Concept Ceragon Proprietary and Confidential Page 10 of 225

11 List of Tables FibeAir IP-10 Series Overview IP-10Q Interfaces Ethernet Interface LEDs Ethernet Interfaces Supported MTU Values Ethernet Interface Functionality Management Interfaces License Types Comparison of IP-10Q Protection Options HSB Protection Switchover Triggers Ethernet Line Protection Comparison Multi-Unit LAG Behavior Header Compression Ethernet Header Compression Comparison Table ACM Working Points (Profiles) IFC and BBS Space Diversity Comparison Automatic State Propagation Port Behavior Example 1 Hybrid Scheduling Example 2 Hierarchical Scheduling IP-10Q Standard and Enhanced QoS Features RFU Selection Guide RFU-C Mechanical, Electrical, and Environmental Specifications RFU-C Mediation Device Losses RFU-C Waveguide Flanges HP/RFU-HP Mechanical, Electrical, and Environmental Specifications HP/RFU-HP Comparison Table New OCB Component Summary All-Indoor Compact Placement Components RFU Models OCB Part Numbers Ceragon Proprietary and Confidential Page 11 of 225

12 OCB Part Numbers for All Indoor Compact All-Indoor Configurations (1+0 /1+1 HSB) All-Indoor Configurations (N+0/N+1 XPIC) All-Indoor Configurations (N+0 / N+1 XPIC Space Diversity) All-Indoor Configurations (N+0 / N+1 XPIC Space Diversity) All-Indoor Configurations (N+0/N+1 Single Pol) All-Indoor Configurations (N+0/N+1 Single Pol Space Diversity) All-Indoor Configurations (N+0/N+1 XPIC Upgrade ready) All-Indoor Configurations (N+0/N+1 XPIC Space Diversity Upgrade-Ready). 159 All-Indoor Configurations (19" Without Rack) RFU-HS Mechanical, Electrical, and Environmental Specifications RFU-SP Frequency Bands RFU-SP Mechanical, Electrical, and Environmental Specifications RFU-HS-SP Antennas P Mechanical, Electrical, and Environmental Specifications P Mediation Device Losses Dedicated Management Ports PolyView Server Receiving Data Ports Web Sending Data Ports Web Receiving Data Ports Additional Management Ports for IP-10Q Supported Ethernet Standards Ceragon Proprietary and Confidential Page 12 of 225

13 About This Guide This document describes the main features, components, and specifications of the FibeAir IP-10Q high capacity IP network solution. This document also describes a number of typical FibeAir IP-10Q configuration options. This document applies to hardware version R1 and software version Q6.9. What You Should Know This document is written for users in North America, and describes applicable standards (ANSI, FCC) for North American users. An ETSI version of this document is also available. Target Audience This manual is intended for use by Ceragon customers, potential customers, and business partners. The purpose of this manual is to provide basic information about the FibeAir IP-10Q for use in system planning, and determining which FibeAir IP-10Q configuration is best suited for a specific network. Related Documents FibeAir IP-10Q Installation Guide - DOC FibeAir IP-10Q User Guide - DOC FibeAir IP-10Q MIB Reference - DOC FibeAir IP-10 License Management System - DOC FibeAir CeraBuild Commission Reports Guide, DOC Ceragon Proprietary and Confidential Page 13 of 225

14 1. Synonyms and Acronyms ACM ACR AES AIS ATPC BBS BER BLSR BPDU BWA CBS CCDP CFM CIR CLI CoS DA DSCP EBS EIR FTP (SFTP) GbE HSB HTTP (HTTPS) IFC IDC IDU LANs LLDP LMS LOF LTE Adaptive Coding and Modulation Adaptive Clock Recovery Advanced Encryption Standard Alarm Indication Signal Automatic Tx Power Control Baseband Switching Bit Error Ratio Bidirectional Line Switch Ring Bridge Protocol Data Units Broadband Wireless Access Committed Burst Size Co-channel dual polarization Connectivity Fault Management Committed Information Rate Command Line Interface Class of Service Destination Address Differentiated Service Code Point Excess Burst Size Excess Information Rate File Transfer Protocol (Secured File Transfer Protocol) Gigabit Ethernet Hot-standby Hypertext Transfer Protocol (Secured HTTP) IF Combining Indoor Controller Indoor unit Local area networks Link Layer Discovery Protocol License Management System Loss Of Frame Long-Term Evolution Ceragon Proprietary and Confidential Page 14 of 225

15 MAID NMS NTP OAM OOF PDV PM PN PSN PTP QoE QoS RDI RFU RMON RSL RSTP SD SFTP SLA SNMP SP STP SSH SSM SyncE TC TOS VC Web EMS WG WFQ WRED WRR XPIC Maintenance Association (MA) Identifier (ID) Network Management System Network Time Protocol Operation Administration & Maintenance (Protocols) Out-of-Frame Packed Delay Variation Performance Monitoring Provider Network (Port) Packet Switched Network Precision Timing-Protocol Quality of-experience Quality of Service Reverse Defect Indication Radio Frequency Unit Ethernet Statistics Received Signal Level Rapid Spanning Tree Protocol Space Diversity Secure FTP Service level agreements Simple Network Management Protocol Strict Priority Spanning Tree Protocol Secured Shell (Protocol) Synchronization Status Messages Synchronous Ethernet Traffic Class Type of Service Virtual Containers Web-Based Element Management System Wave guide Weighted Fair Queue Weighted Random Early Detection Weighted Round Robin Cross Polarization Interference Cancellation Ceragon Proprietary and Confidential Page 15 of 225

16 2. Introduction This chapter includes: Product Overview IP-10Q Advantages Functional Description Single Point of Management Solution Overview Ceragon Proprietary and Confidential Page 16 of 225

17 2.1 Product Overview FibeAir IP-10Q is a high capacity carrier-grade wireless Ethernet backhaul product. IP-10Q is designed for high capacity (full GbE) aggregation wireless links, with an ultra high-density design that enables the use of four carriers in a single rack unit (1RU). FibeAir IP-10Q covers the entire licensed frequency spectrum and offers a wide capacity range, from 50 Mbps to 1 Gbps over a single radio carrier, using a single Radio Frequency Unit (RFU), depending on traffic scenario based on MAC and enhanced Multi-Layer header compression. Additional functionality and capacity are enabled via license keys while using the same hardware. By enabling more capacity, at lower latencies to any location, with proper traffic management mechanisms and an optional downstream boost, FibeAir IP-10Q is built to enhance end user Quality of Experience. Highlights of IP-10Q include: Best utilization of spectrum assets Improved network uptime Future proof Risk-free solution Ceragon Proprietary and Confidential Page 17 of 225

18 2.2 IP-10Q Advantages The following are just some of the advantages that IP-10Q provides Efficient Utilization of Spectrum Assets IP-10Q provides efficiencies at three levels -- spectral efficiency, radio link, and wireless network. By combining superior radio performance, advanced compression, and a holistic end-to-end approach to capacity, operators can effectively provide up to five times more traffic to their users. In other words, IP-10Q enables more revenue-generating subscribers in a given RAN Spectral Efficiency IP-10Q provides a high degree of spectral efficiency in a given spectrum channel by optimizing link capacity using adaptive coding and modulation techniques. In addition, IP-10Q provides several options for header compression: Legacy MAC header compression Provides up to 45% in additional Ethernet throughput. Multi-Layer (Enhanced) header compression (license-enabled) Provides up to 300% additional effective Ethernet throughput, depending on frame size, channel bandwidth, and modulation Radio Link Latency IP-10Q boasts ultra-low latency features that are essential for 3G and LTE deployments. With low latency, IP-10Q enables links to cascade further away from the fiber PoP, allowing wider coverage in a given network cluster. Ultra-low latency also translates into longer radio chains, broader radio rings, and shorter recovery times. Moreover, maintaining low packet delay variation ensures proper synchronization propagation across the network. System Gain IP-10Q s high system gain enables the use of small antennas and long link spans, resulting in high overall capacity while maintaining critical and real-time traffic, saving on both operational and capital expenditures by using smaller antennas for a given link budget Wireless Network Enhanced QoS IP-10Q enables operators to deploy differentiated services with stringent service level agreements while maximizing the utilization of network resources. IP-10Q enables granular CoS classification and traffic management, network utilization monitoring, and support of EIR traffic without affecting CIR traffic. Enhanced QoS enables traffic shaping per queue and port in order to limit and control packet bursts, and improves the utilization of TCP flows using WRED protocols. Ceragon Proprietary and Confidential Page 18 of 225

19 OA&M With advanced service management and Operation Administration & Maintenance (OA&M) tools, IP-10Q simplifies network design, reduces operational and capital expenditures, and improves overall network availability and reliability to support services with stringent SLA Scalability FibeAir IP-10Q is a scalable solution that is based on a common hardware that supports any channel size, modulation scheme, capacity, network topology, and configuration. Scalability and hardware efficiency simplify logistics and allow for commonality of spare parts. A common hardware platform enables customers to upgrade the feature set as the need arises - Pay As You Grow - without requiring hardware replacement Availability MTBF. FibeAir IP-10Q provides an unrivaled reliability benchmark, with radio MTBF exceeding 112 years, and average annual return rate around 1%. Ceragon radios are designed in-house and employ cutting-edge technology with unmatched production yield, and a mature installed-base exceeding 100,000 radios. In addition, advanced radio features such as multi-radio and cross polarization (XPIC) enable the system to achieve 100% utilization of radio resources by load balancing based on instantaneous capacity per carrier. Important resulting advantages are reduction in capital expenditures due to less spare parts required for rollout, and reduction in operating expenditures, since maintenance and troubleshooting are infrequently required. ACM Adaptive Modulation has a remarkable synergy with FibeAir IP- 10Q s built-in Layer 2 QoS mechanism. Since QoS provides priority support for different classes of service, according to a wide range of criteria, it is possible to configure the system to discard only low priority packets as conditions deteriorate. Adaptive Power and Adaptive Coding & Modulation provides maximum availability and spectral efficiency in any deployment scenario Network Level Optimization FibeAir IP-10Q optimizes overall network performance, scalability, resilience, and survivability by using hot-standby (HSB) configurations with no single point of failure. FibeAir IP-10Q helps create a more robust network, with minimum downtime and maximum service grade, ensuring better user experience, better immunity to failures, lower churn, and reduced expenditures. Ceragon Proprietary and Confidential Page 19 of 225

20 2.2.8 Network Management Each IP-10 Network Element includes an HTTP web-based element management system (Web EMS) that enables the operator to perform element configuration, RF, Ethernet, and PDH performance monitoring, remote diagnostics, alarm reports, and more. In addition, FibeAir IP-10Q provides an SNMP-based northbound interface for network management. For network management, Ceragon offers NetMaster, a comprehensive NMS that provides centralized operation and maintenance capability for the complete range of network elements in an IP-10Q system. NetMaster is built using state-of-the-art technology as a scalable, cross-platform NMS that supports distributed network architecture. Ceragon also offers PolyView, with best-in-class end-to-end Ethernet service management, network monitoring, and NMS survivability using advanced OAM. PolyView provides simplified network provisioning, configuration error prevention, monitoring, and troubleshooting tools that ensure better user experience, minimal network downtime, and reduced expenditures on network-level maintenance Power Saving Mode with High Power Radio FibeAir IP-10Q offers an optional ultra-high power radio solution that transmits the highest power in the industry, while employing an innovative Power Saving Mode that saves up to 30% power consumption. Power Saving Mode enables the deployment of smaller antennas, and reduces the need for repeater stations. Moreover, installation labor cost and electricity consumption are reduced, achieving an overall diminished carbon footprint. Ceragon Proprietary and Confidential Page 20 of 225

21 2.3 Functional Description Featuring an advanced architecture, FibeAir IP-10Q uniquely integrates the latest radio technology with Ethernet networking. The IP-10Q radio core engine is designed to support native Ethernet over the air interface enhanced with Adaptive Power and Adaptive Coding & Modulation (ACM) for maximum spectral efficiency in any deployment scenario. The modular design is easily scalable with the addition of units or license keys. Functional Block Diagram Functional Overview Up to four FibeAir IP-10Q IDUs can be installed in a single chassis. An IP-10Q chassis fits in a single Rack Unit (1RU). The chassis has a backplane, which provides connectivity among the modules in the chassis for XPIC, Multi-Radio, BBS Space Diversity, and management. The chassis backplane enables unified management of the system as a single network element with multiple radio links. Refer to Single Point of Management on page 23. The IP-10Q chassis also provides an inherent dual-feed power solution by which the power feed for each module on a single floor of the chassis can feed power to the other module on the same floor of the chassis. This provides power redundancy in any configuration that includes two modules on a single floor of the chassis. For example, in a 2 x 2+0 configuration, both Module 1 and Module 2, and Module 3 and Module 4 provide each other with power feed redundancy. IP-10Q has a DC input voltage nominal rating of -48V. Ceragon Proprietary and Confidential Page 21 of 225

22 FibeAir IP-10Q Block Diagram The CPU acts as the module s central controller, and all management frames received from or sent to external management applications must pass through the CPU. In a chassis that contains more than one module, the main module s CPU serves as the central controller for the entire chassis. The Mux assembles the radio frames, and holds the logic for protection and Space Diversity. The modem represents the physical layer, modulating, transmitting, and receiving the data stream. The following figure shows how modules in an IP-10Q chassis are numbered. Module 1, on the lower left, is the main module for purposes of unified chassis management. In a protected 1+1 configuration, Module 2 functions as a backup for Module 1. Ceragon Proprietary and Confidential Page 22 of 225

23 IP-10Q Chassis - Module Numbering Single Point of Management In a chassis with multiple IDUs, all management is performed through the main module. The main module communicates with the other modules in the chassis through the chassis backplane. The main module s CPU operates as the central controller for the chassis, and all management frames received from or sent to external management applications must pass through the CPU. An IP-10Q chassis has a single IP management address, which is the address of the main module. In a protected 1+1 configuration, the chassis has two IP addresses, those of each of the main modules. The IP address of the active main module is used to manage all the modules in the chassis. Several methods can be used for IP-10Q chassis management: Local terminal CLI CLI via telnet Web-based management SNMP The NMS represents the chassis as a single unit. The Web-Based EMS enables access to all IDUs in the chassis from its main window. In addition, the management system provides access to other network equipment through in-band or out-of-band network management. To ease the reading and analysis of several IDU alarms and logs, the system time should be synchronized to the main module s time Centralized System Features The following IP-10Q functions are configured centrally through the main module in the chassis: IP Communications All communication channels are opened through the main module s IP address. User Management Login, adding users, and deleting users are performed centrally. Chassis Time Synchronization System time is automatically synchronized for all IDUs in the chassis. Chassis Software Version Management Software can be upgraded or downgraded in all IDUs in the chassis from the main module. Ceragon Proprietary and Confidential Page 23 of 225

24 Chassis Configuration Backup Configuration files can be created, downloaded, and uploaded from the main module. Chassis Reset Modules in the chassis can be reset individually or collectively both from the main module and locally. All other functions are performed for each IDU individually. Ceragon Proprietary and Confidential Page 24 of 225

25 2.4 Solution Overview Single Carrier/Single Direction IP-10Q is part of the FibeAir IP-10 series that includes IP-10G, packet-only IP- 10E, all-outdoor IP-10C for access, and high-capacity high-density IP-10Q, which is optimized for high-capacity MPLS-aware Ethernet microwave radio where fiber connections are not available. The FibeAir series provides a variety of solutions for a large number of deployment scenarios. TDM and Ethernet FibeAir IP-10 Series Overview Ethernet IP-10G IP-10E IP-10C Multi-Carrier/Multi Direction Integrated Backhaul (L2) Smart Pipe (L1) IP-10G Nodal IP-10E Nodal IP-10Q Ceragon Proprietary and Confidential Page 25 of 225

26 3. Hardware Description This chapter includes: Hardware Architecture Front Panel Description Ethernet Interfaces Management Interfaces Radio Interface Power Interfaces Additional Interfaces Front Panel LEDs External Alarms Ceragon Proprietary and Confidential Page 26 of 225

27 3.1 Hardware Architecture FibeAir IP-10Q features split-mount architecture consisting of an indoor unit (IDU) and a Radio Frequency Unit (RFU). An IF cable connects the IDU to the RFU, transmits traffic and management data between the IDU and the RFU, and provides 48 V power to the RFU. IDUs are placed in a chassis that fits into a single Rack Unit (1RU) and contains up to four IP-10Q IDUs. The IP-10Q chassis has two floors, with room for two IDUs on each floor. When two IDUs are installed on a floor of the chassis, the power feed for each IDU provides a backup power source for the other IDU on the floor by means of the chassis backplane. The chassis backplane also provides interfaces for management, Multi-Radio, protection, and XPIC. Each IP-10Q unit includes the following interfaces. Main Interfaces: 1 x GbE combo port for traffic: 10/100/1000Base-T or SFP 1000Base-X 2 x GbE ports for management: 10/100/1000Base-T RFU interface: TNC connector Additional Interfaces: Terminal console External alarms (4 inputs and 1 output) IP-10Q can work with a variety of RFU types, including split-mount, remotemount, and all-indoor configurations. A description of each RFU, as well as a comparison chart of the capacity and features supported in each RFU, is provided in this document. Available assembly options are: With or without XPIC support For additional information: Radio Frequency Units Ceragon Proprietary and Confidential Page 27 of 225

28 3.2 Front Panel Description This section describes the IP-10Q s front panel. The following sections provide detailed descriptions of the IP-10Q interfaces. IP-10Q Front Panel and Interfaces IP-10Q Interfaces Interface For Further Information 1 x GbE combo port for traffic Ethernet Interfaces 2 x GbE ports for management Ethernet Interfaces Craft Terminal RFU Interface Power Interface Front Panel Alarms Additional Interfaces Radio Interface Power Interfaces Front Panel LEDs Ceragon Proprietary and Confidential Page 28 of 225

29 3.3 Ethernet Interfaces Related Topics: Multi-Unit LAG FibeAir IP-10Q has a GbE Ethernet interface for traffic and two GbE interfaces for management on the front panel. For the traffic interface, you can choose between an optical and an electrical physical interface. Each traffic interface is labeled 1. The optical interface is located to the right of the electrical interface. The management interfaces are labeled 2 and 3, and are located to the right of the traffic interfaces. The following table describes the Ethernet interface LEDs. Ethernet Interface LEDs Interface Functionality LED (right) Activity LED (left) Combo Eth1 (RJ-45) Combo Eth1 (SFP) Eth2 Eth3 When the port is enabled and interface type is electrical RJ-45, the LED will be on. Otherwise it will be off. The SFP LED (below the SFP interface) will be on when the port is enabled and a carrier is detected. This LED will blink when traffic passes. When the port is enabled and used for management, the LED will be on. When the port is enabled and used for management, the LED will be on. When a carrier is detected, the LED will be on. When traffic passes, the LED will blink. Disabled When a carrier is detected, the LED will be on. When traffic passes, the LED will blink. When a carrier is detected, the LED will be on. When traffic passes, the LED will blink. The following table shows the MTU values supported by the IP-10Q Ethernet interfaces. Ethernet Interfaces Supported MTU Values Interface type Jumbo mode Non jumbo mode Ethernet Traffic port MTU = 9612 MTU = 1632 Management port MTU = 1632 MTU = 1632 Note: In non jumbo mode, the RMON oversized frames counter will count frames that exceed 2048 bytes. In jumbo mode, the RMON oversized frames counter will only count frames that exceed bytes. It is possible to use an electrical interface at one end of the link, and an optical interface at the other end. In order to change interfaces, it is essential to disable the active interface first, and then to enable the other interface. Ceragon Proprietary and Confidential Page 29 of 225

30 Ethernet Interface Functionality Interface Name Interface Rate Functionality Eth1 (leftmost dual SFP/RJ-45) Electrical GbE - 10/100/1000 OR Optical GbE Traffic Eth2 Eth3 Eth4 (internal) Electrical GbE - 10/100/1000 Management only Electrical GbE - 10/100/1000 Management only Electrical GbE - 10/100/1000 Traffic or Protection mirroring Eth5 (Radio) According to Radio script Traffic Note: Internal GbE port 4 can be used as a user interface port instead of GbE port 1. To provide this functionality, the user must first disable port 1 and then enable port 4. Management ports Eth2 and Eth3 can be also GbE according to user configuration. When HSB protection and Multi Unit LAG are enabled, port4 is used as a protection mirroring port GbE Interfaces The IP-10Q supports two dual GbE interface, which the user can configure to Electrical GbE (10/100/1000) or Optical GbE (SFP) interface. It is NOT supported and NOT possible to use SFP with electrical stack. SFP supports only optical stack. It is possible to use an electrical interface at one end of the link, and an optical interface at the other end. In order to change interfaces, it is essential to disable the active interface first, and then to enable the other interface. The options are: Eth1: Electrical GbE or Optical GbE. Eth4: Internal Electrical GbE Ethernet port4 is connected on the backplane to port4 of the mate unit on the same floor of the shelf. This port perfectly suits for east-west and pipe chain applications. Ceragon Proprietary and Confidential Page 30 of 225

31 The following table lists recommended SFP manufacturers. Part Number Item Description Manufacturer Name Manufacturer PN AO XCVR,SFP,850nm,1.25Gb,MM,500M,W.DDM PHOTON PST120-51TP+ AO XCVR,SFP,850nm,1.25Gb,MM,500M,W.DDM Wuhan Telecom. Devices (WTD) RTXM AO XCVR,SFP,850nm,1.25Gb,MM,500M,W.DDM CORETEK (*) CT-1250NSP-SB1L AO XCVR,SFP,850nm,1.25Gb,MM,500M,W.DDM Fiberxon FTM-8012C-SLG AO XCVR,SFP,1310nm,1.25Gb,SM,10km Wuhan Telecom. Devices (WTD) RTXM AO XCVR,SFP,1310nm,1.25Gb,SM,10km CORETEK (*) CT-1250TSP-MB4L-A AO XCVR,SFP,1310nm,1.25Gb,SM,10km Fiberxon FTM-3012C-SLG AO XCVR,SFP,1310nm,1.25Gb,SM,10km AGILENT AFCT-5710PZ * Electrically, these SFP modules work properly but they tend to get mechanically stuck in the IP-10 cage. Ceragon Proprietary and Confidential Page 31 of 225

32 3.4 Management Interfaces An IP-10Q system can be configured to use 1 or 2 Ethernet management ports. Interfaces Eth2 and Eth3 are the only interfaces that can be assigned as management ports. Management Interfaces Configured Number of Management Ports 1 Eth3 2 Eth3, Eth2 0 None Management Interfaces Management interfaces are connected to the switch (bridge) and are configured to learning mode. Management frames should always be assigned maximum priority in order to ensure that network management remains available in a loaded network. In order to achieve this, the IP-10Q automatically assigns to all management frames (frames incoming from the management interfaces) a p-bit value of 7, which is the highest priority by default. Management interfaces can be configured to have one of the following capacities: 64kbps, 128kbps, 256kbps, 512kbps, 1024kbps, 2048kbps (default). Capacity is limited by the port ingress rate limit. Ceragon Proprietary and Confidential Page 32 of 225

33 3.5 Radio Interface The IP-10Q s radio interface is represented in the system as Eth5. The radio interface uses a TNC connector to connect, via a coaxial cable, to the RFU. Ceragon Proprietary and Confidential Page 33 of 225

34 3.6 Power Interfaces The IP-10Q power supply is protected by dual power feed mechanism with hitless power source switching. Each IDU can be fed from its own power connector or from a connector of the mate IDU. Note: The mate IDU is the IDU on the same floor of the rack. For example, the IDUs in slots 1 and 2 are mates for each other. The fans unit can receive power from the IDU in slot 1 or the IDU in slot 2. Ceragon Proprietary and Confidential Page 34 of 225

35 3.7 Additional Interfaces An IP-10Q contains the following additional interfaces: Terminal Console The terminal console is an RJ-45 interface. A local craft terminal can be connected to the terminal console for local CLI management of the individual IDU. If the IDU is the main module, access to other units in the chassis is also available through the terminal console of the main module. The terminal console has the following parameters: Baud: Data bits: 8 Parity: None Stop bits: 1 Flow Control: None Engineering Order Wire (EOW) (optional) Ceragon Proprietary and Confidential Page 35 of 225

36 3.8 Front Panel LEDs The following LEDs are located beneath the external alarms on the front panel: LINK Indicates status of the radio link. IDU Indicates status of the Ethernet interface. RFU Indicates status of the RF module. PROT Indicates the main and standby unit alarm and protection status. RMT Indicates status of the remote unit. These LEDs indicate the following: LINK IDU RFU PROT RMT Green Radio link is operational Orange Minor BER alarm on the radio Red Loss of signal, major BER alarm on the radio Green IDU is functioning normally Orange Fan failure Red Alarm on IDU (all severities) Green RFU is functioning normally Orange Loss of communication between the IDU and the RFU Red RFU failure Main Unit Green No alarms Standby Unit Yellow No alarms Orange Forced switch, protection lock Red Physical errors (no cable, cable failure) Off Protection is disabled, or not supported on the device Green Remote IDU is functioning normally Orange Minor alarm on the remote IDU Red Major alarm on the remote IDU Ceragon Proprietary and Confidential Page 36 of 225

37 3.9 External Alarms IP-10Q includes a DB9 dry contact external alarms interface. The external alarms interface supports five input alarms and a single output alarm. The input alarms are configurable according to: 1 Intermediate 2 Critical 3 Major 4 Minor 5 Warning The output alarm is configured according to predefined categories. Ceragon Proprietary and Confidential Page 37 of 225

38 4. Licensing This chapter includes: License Overview Working with License Keys Licensed Features Ceragon Proprietary and Confidential Page 38 of 225

39 4.1 License Overview FibeAir IP-10Q offers a pay as-you-grow concept to reduce network costs. Future capacity growth and additional functionality is enabled with license keys using the same hardware. Licenses are divided into two categories: Per Radio Each IDU (both sides of the link) require a license. Per Configuration Only one license is required for the system. A 1+1 configuration requires the same set of licenses for both the active and the protected IDU. For licenses that are not per radio, licenses should be assigned to the main (bottom, left) IDU in the chassis. 4.2 Working with License Keys Ceragon provides a web-based License Management System (LMS). The LMS enables authorized users to generate license keys, which are generated per IDU serial number. In order to upgrade a license, the license-key must be entered into the IP-10Q, followed by a cold reset. When the system returns online following the reset, its license key is checked and implemented, enabling access to new capacities and/or features. For more detailed information, refer to FibeAir IP-10 License Management System, DOC Licensed Features As your network expands and additional functionality is desired, license keys can be purchased for the features described in the following table. License Types License Name Description For Addition Information Adaptive Coding and Modulation (ACM) Capacity Upgrade Synchronization Unit Enables the Adaptive Coding and Modulation (ACM) feature. An ACM license is required per radio. If additional IDUs are required for non-radio functionality, no license is required for these units. Enables you to increase your system s radio capacity in gradual steps by upgrading your capacity license. Enables the Synchronization unit required for SyncE support. Adaptive Coding Modulation (ACM) Synchronization Ceragon Proprietary and Confidential Page 39 of 225

40 License Name Description For Addition Information Enhanced QoS Enhanced Header Compression Enables the Enhanced QoS feature, which includes eight priority queues with configurable buffer length, a larger selection of classification criteria, WRED for improved congestion management, an enhanced scheduler based on Strict Priority, Weighted Fair Queue (WFQ), or a hybrid approach that combines Strict Priority and WFQ, and other enhanced functionality. A license is required per radio. Enables the use of Multi-Layer header compression, which can increase effective throughput by up to 300%. Enhanced QoS Ethernet Header Compression Ceragon Proprietary and Confidential Page 40 of 225

41 5. Feature Description This chapter includes: Equipment Protection Ethernet Line Protection Capacity and Latency Radio Features Ethernet Features Quality of Service (Traffic Manager) Synchronization Ceragon Proprietary and Confidential Page 41 of 225

42 5.1 Equipment Protection This section includes: Equipment Protection Overview 1+1 HSB Protection 2+0 Multi-Radio and 2+0 Multi-Radio with IDU and Line Protection Switchover Triggers Related topics: Ethernet Line Protection Floating IP Address Ceragon Proprietary and Confidential Page 42 of 225

43 5.1.1 Equipment Protection Overview IP-10Q offers several options for line and equipment protection. The following protected configurations are available: 1+1 HSB (with optional Space Diversity using BBS) 2+0 Multi-Radio 2+0 Multi-Radio with IDU and Line Protection The following table summarizes the degree of protection provided by the various IP-10Q configuration options: Comparison of IP-10Q Protection Options Configuration # of IDUs per Terminal # of RFUs per Terminal Radio Capacity Normal Radio Capacity Unit Failure XPIC Support ACM Support BBS SD Support No Optional No 1+1 HSB No Optional 1 Optional 2+0 Multi-Radio RFU Failure 1 2 IDU (Slave) Failure 1 3 IDU (Master) Failure - 0 Optional Optional No 2+0 Multi-Radio with IDU and Line Protection RFU Failure 1 4 IDU (Slave or Master) Failure Optional Optional No ACM is not supported when BBS Space Diversity is used. With graceful degradation. With graceful degradation. With graceful degradation. With graceful degradation. Ceragon Proprietary and Confidential Page 43 of 225

44 HSB Protection This feature cannot be used with the following: Multi-Radio 2+0 Multi-Radio with IDU and line protection Related topics: Adaptive Coding Modulation (ACM) A 1+1 configuration scheme can be used to provide full protection in the event of IDU or RFU failure. The two IDUs operate in active and standby mode. If there is a failure in the active IDU or RFU, the standby IDU and RFU pair switches to active mode. In a 1+1 configuration, the active IDU must be placed in the lower left slot of the chassis (Module 1). The standby IDU must be placed in the lower right slot of the chassis (Module 2). The IDUs are connected by the backplane of the chassis. 1+1 HSB can be used with BBS Space Diversity. The following figure illustrates a 1+1 HSB Space Diversity configuration. 1+1 HSB Node with BBS Space Diversity IP-10Q units in a 1+1 HSB configuration constitute a completely redundant system, including management. Each unit can be managed with its own IP address, and the whole chassis can be accessed via the active unit. To ensure that the user can always access the active unit directly, even in the event of switchover, a floating IP address can be configured. This provides a single IP address that will always provide direct access to the currently active main unit. In a 1+1 HSB configuration, it is necessary for both units to have the same configuration. IP-10Q includes a mismatch mechanism that detects if there is a mismatch between the configurations of the local and mate units. This mechanism is activated by the system periodically and independently of other protection mechanisms, at fixed intervals. It is activated asynchronously in both the active and the standby units. Once the mismatch mechanism detects a configuration mismatch, it raises a Mate Configuration Mismatch alarm. When the configuration of the active and standby unit is changed to be identical, the mechanism clears the Mate Configuration Mismatch alarm. For addition information: Switchover Triggers Floating IP Address Ceragon Proprietary and Confidential Page 44 of 225

45 Multi-Radio and 2+0 Multi-Radio with IDU and Line Protection This feature cannot be used with the following: 1+1 HSB Space diversity ACM Related topics: Multi-Radio 2+0 Multi-Radio provides a significant degree of protection, in addition to doubling capacity by enabling two separate radio carriers to be shared by a single Ethernet port. In the event of RFU failure, or failure of the slave IDU, one RFU and IDU remain in operation, with graceful degradation of service to ensure that not all data is lost, but rather, a reduction of bandwidth occurs. However, if there is a failure of the master IDU, traffic and management access is lost. The IDU and line protection option increases protection to the master IDU. If there is a failure in the master IDU, the slave IDU becomes the master, and continues to provide service. Thus, a 2+0 Multi-Radio configuration with IDU and line protection provides protection for the failure of any IDU or RFU in the node. The IDU and line protection feature protects Ethernet traffic. It also protects management of the chassis, since chassis management is handled by the master IDU. Graceful degradation is provided with the help of IP-10Q s integrated QoS mechanism, which ensures that high-priority traffic is maintained in the event of reduced bandwidth Multi-Radio with IDU and Line Protection Basic Operation Multi-Radio with line protection is available for adjacent pairs of IDUs (slots 1 and 2 or 3 and 4). The active unit is the IDU that currently holds the line interfaces and it is also a Multi-Radio master unit. The following diagram illustrates the traffic flow in Multi-Radio with line protection. Ceragon Proprietary and Confidential Page 45 of 225

46 Multi-Radio 2+0 with Line Protection Traffic Flow Ethernet Ethernet Ethernet Ethernet Ethernet Ethernet Ethernet TDM Ethernet Ethernet Orange lines represent the Ethernet traffic flow. The active IDU holds the line interfaces for Ethernet traffic. The active IDU acts as a Multi-Radio master unit by distributing the Ethernet traffic between its own radio channel and the radio channel of its mate. At the receive side of the link, the active IDU combines the data from both radio channels to create a single Ethernet stream. When a protection switch occurs, the new active IDU also becomes the Multi-Radio main unit. The following events will cause a protection switchover: Note: GbE line Loss of Carrier (LOC) User manual switch For addition information: Switchover Triggers Radio failure or BER in the radio channel will not cause a protection switchover. Multi-Radio protects against radio channel failure by blocking the defective radio. Ceragon Proprietary and Confidential Page 46 of 225

47 5.1.4 Switchover Triggers Switchover triggers for 1+1 HSB protection configurations are described in the following table, according to their priority, with the highest priority triggers on top. HSB Protection Switchover Triggers Priority Fault Remark 1 Mate Power OFF - 2 Lockout Does not persist after cold reset. 3 Force Switch Does not persist after cold reset. 4 Local Radio LOF - 5 SFP LOS/GBE LOC Electrical GBE LOC is configurable. Only the active unit is monitored in this case. 6 Change Remote request due to "Radio LOF" - 7 Local Radio Excessive BER Configurable. Irrelevant in ACM adaptive mode 8 Change Remote due to Radio Excessive BER Irrelevant in ACM adaptive mode 9 Manual Switch - Ceragon Proprietary and Confidential Page 47 of 225

48 5.2 Ethernet Line Protection This section includes: Ethernet Line Protection Options Multi-Unit LAG Ceragon Proprietary and Confidential Page 48 of 225

49 5.2.1 Ethernet Line Protection Options IP-10Q offers a number of Ethernet line protection options for various multiunit configuration scenarios in which two IP-10Q IDUs are connected to an external switch or router. These are: Single Interface with Splitter A single interface in the external switch or router is connected to each of the two IDUs using a splitter. A splitter can be used with the optical GbE port on each IDU. Dual Interface with Multi-Unit LAG Two interfaces in the external switch or router are configured as a static LAG, and each interface is connected to one IDU. Full protection of each interface is provided by a LAG that includes interfaces in both IDUs. Multi-Unit LAG can be used with the electrical GbE ports. Hardware Protection with Single Interface Using Optical Splitter Full Protection Using Multi-Unit LAG All of these line protection methods are available for any of the following configurations: 1+1 HSB 2+0 Multi Radio with IDU and Line Protection All BBS space diversity configurations Ceragon Proprietary and Confidential Page 49 of 225

50 The following table compares the advantages and limitations of the Ethernet line protection schemes described in this section. Ethernet Line Protection Comparison Protection Scheme Extent of Protection Interfaces Splitters Required Single Interface with Optical Splitter Dual interface with Multi-Unit LAG Protection for failure of IDU interface, but not for failure of external switch/router interface. Full Ethernet line protection for IDU and switch/router interfaces. Optical GbE Yes Electrical GbE No Ceragon Proprietary and Confidential Page 50 of 225

51 5.2.2 Multi-Unit LAG With Multi-Unit LAG, the switch or router relates to the IDUs as a single device. There is no need for splitters, and Multi-Unit LAG can be used to protect either the electrical GbE port or the optical GbE port. 6 In contrast, splitters can only be used to protect the optical GbE port. The service disruption time in case of failure in one of the LAG physical ports is less than 50ms in most cases using Multi-Unit LAG. An IP-10Q system using Multi-Unit LAG has dual (redundant) GbE interfaces. Each of these interfaces is connected to a separate interface on an external switch or router. The IP-10Q interfaces are active and enabled on both the active or master unit and the standby or slave unit. On the external unit, a static LAG must be configured on the interfaces that are connected to the IDUs. If the IP-10Q IDUs are in Multi-Radio mode with IDU and line protection, any link failure triggers graceful degradation and is transparent to the external unit. If an IDU itself experiences unit failure, the interface to which it is connected on the external unit is disabled. If the disabled IDU is the standby unit, or if it is the active unit and Multi-Radio with IDU and line protection is enabled, the functioning IDU maintains connectivity with the external unit via the interface to which the functioning IDU is connected. Multi-Unit LAG is supported with any of the following protection features: 1+1 HSB 1+1 Space Diversity 2+0 Multi Radio with line protection Multi-Unit LAG supports both electrical and optical interfaces. The following figure illustrates the basic operation of Multi-Unit LAG. Multi-Unit LAG Basic Operation 6 In software release Q6.9, Multi-Unit LAG is only supported for electrical ports. Ceragon Proprietary and Confidential Page 51 of 225

52 An external switch is connected to the HSB-protected IP-10Q link by means of two static Link Aggregation (LAG) ports. The external switch can be another IP-10Q IDU or any third party equipment that supports static LAG protocol. The first LAG port of the external switch is connected to Ethernet port 1 of the active IP-10Q unit and the second LAG port is connected to Eth1 of the standby IP-10Q unit. Internal Eth4 is used for traffic mirroring, as described below. In the uplink direction (toward the radio), the external switch splits the packets between the two LAG interfaces, which are connected to the active and standby IP-10Q units. Ethernet packets received from the LAG interface in the active IP-10Q unit are sent to the radio. Ethernet packets received from the LAG interface in the standby IP-10Q unit are mirrored to the active IP-10Q unit on Eth4. The active unit receives these packets from Eth4 and sends them to the radio. In the downlink (from the radio), the active IP-10Q unit receives Ethernet packets from the radio and forwards all of the packets to the External Switch through Eth1. The following table describes the behavior of Multi-Unit LAG Ethernet line protection. Multi-Unit LAG Behavior Scenario Failure in Eth1 in active IDU Failure in Eth1 in standby IDU Failure in the mirroring port Reaction Initiate protection switchover. LAG protocol on the external switch recognizes the port failure and uses the second LAG port (the one that is connected to the active IP-10Q unit). No protection switchover is initiated. Standby unit shuts down port1 to indicate failure to the external switch. After resolving the failure, the standby unit reopens port1 automatically. No protection switchover is initiated. Note: The mirroring port failure is very rare since this port is internal. Such a failure may be caused by a damaged back connector of the IDU or by improper insertion of the IDU into the chassis. Ceragon Proprietary and Confidential Page 52 of 225

53 5.3 Capacity and Latency This section includes: Capacity Summary Ethernet Header Compression Latency Ceragon Proprietary and Confidential Page 53 of 225

54 5.3.1 Capacity Summary Modulations QPSK to 256 QAM Radio capacity Up to 251/324/428/465 Mbps throughput over 30/40/50/80 MHz channels Radio capacity with legacy MAC Header Compression Up to 287/370/489/532 Mbps throughput Radio capacity with Multi-Layer (Enhanced) Header Compression (license-enabled) 728/938/1,000/1000 Mbps throughput. All licensed bands L6, U6, 7, 8, 10, 11, 13, 15, 18, 23, 26, 28, 31, 38 GHz Highest scalability From 50 Mbps to 500 Mbps, using the same hardware, including the same RFU, and up to 1 Gbps with Multi-Layer Enhanced Header Compression. For additional information: Radio Capacity Specifications Ceragon Proprietary and Confidential Page 54 of 225

55 5.3.2 Ethernet Header Compression IP-10Q offers several Ethernet header compression methods, which enable operators to significantly improve Ethernet throughout over the radio link without affecting user traffic: No Header Compression (Layer 1 Header Suppression) Removes the IFG and Preamble fields. This mechanism operates automatically even if no header compression is selected by the user. MAC Header Compression ( Legacy Mode ) Operates at Layer 2, compressing the MAC SA and the MAC DA. The user can enable or disable MAC header compression. Multi-Layer Header Compression ( Enhanced Compression ) Users can configure the depth of Enhanced Compression, up to Layer 4. Enhanced Compression requires software version q6.9. Enhanced Compression also requires a license. Header Compression Ceragon Proprietary and Confidential Page 55 of 225

56 Layer 1 Header Suppression Even when no header compression is enabled, IP-10Q performs Layer 1 header suppression. Layer 1 header suppression removes the IFG and Preamble fields (20 bytes), replacing them with a GFP header. Headers fields in Layers 2 through 4 are not compressed at all. The following figure provides a detailed diagram of Layer 1 header suppression. Layer 1 Header Suppression 12B 8B Inter-Frame Gap (IFG) Preabmle L1 header (PHY) 6B MAC DA 4B 6B 6B 2B 2B 2B 2B 2B 4B GFP header MAC DA MAC SA 0x8A88 (opt) S-Vlan (opt) 0x8100 (opt) C-Vlan (opt) 0x0800/0x86DD L3/L4 headers (optional) & Payload CRC 6B 2B 2B 2B 2B 2B 4B MAC SA 0x8A88 (opt) S-Vlan (opt) 0x8100 (opt) C-Vlan (opt) 0x0800/0x86DD L3/L4 headers (optional) & Payload CRC L2 header (MAC) MAC Ceragon Proprietary and Confidential Page 56 of 225

57 MAC Header Compression ( Legacy Mode ) IP-10Q s legacy MAC header compression operates on Layer 2, and supports up to eight flows. Legacy MAC header compression improves effective throughput over the radio link by up to 45% or more without affecting user traffic. Legacy MAC header compression compresses the MAC SA and the MAC DA fields (12 bytes). Layer 1 header suppression is also active, replacing the IFG and Preamble fields (20 bytes) with a GFP header. Legacy MAC header compression does not require a license, and can be enabled and disabled by the user. By default, legacy MAC header compression is disabled. The following figure provides a detailed diagram of how the frame structure is affected by legacy MAC header compression. Legacy MAC Header Compression 12B 8B Inter-Frame Gap (IFG) Preabmle L1 header (PHY) 4B 1B 2B 2B 2B 2B 2B 4B GFP header Flow ID 0x8A88 (opt) S-Vlan (opt) 0x8100 (opt) C-Vlan (opt) 0x0800/0x86DD L3/L4 headers (optional) & Payload CRC 6B 6B 2B 2B 2B 2B 2B 4B MAC DA MAC SA 0x8A88 (opt) S-Vlan (opt) 0x8100 (opt) C-Vlan (opt) 0x0800/0x86DD L3/L4 headers (optional) & Payload CRC L2 header (MAC) MAC Ceragon Proprietary and Confidential Page 57 of 225

58 Multi-Layer (Enhanced) Header Compression This feature requires: Enhanced Header Compression license Related topics: Licensing Multi-Layer (Enhanced) header compression identifies traffic flows and replaces the header fields with a "flow ID". This is done using a sophisticated algorithm that learns unique flows by looking for repeating frame headers in the traffic stream over the radio link and compressing them. The principle underlying this feature is that packet headers in today s networks use a long protocol stack that contains a significant amount of redundant information. In Enhanced Compression mode, the user can determine the depth to which the compression mechanism operates, from Layer 2 to Layer 4. Operators must balance the depth of compression against the number of flows in order to ensure maximum efficiency. Up to 256 concurrent flows are supported. Up to 68 bytes of the L2-4 header can be compressed. In addition Layer 1 header suppression is also performed, replacing the IFG and Preamble fields (20 bytes) with a GFP header. Multi layer header compression can be used to compress the following types of header stacks: Ethernet MAC untagged IPv4 IPv6 TCP UDP TCP MPLS UDP Ethernet MAC + VLAN IPv4 IPv6 TCP UDP TCP MPLS UDP Ethernet MAC with QinQ IPv4 IPv6 TCP UDP TCP Ceragon Proprietary and Confidential Page 58 of 225

59 MPLS PBB-TE UDP The following figure provides a detailed diagram of how the frame structure is affected by Multi-Layer (Enhanced) header compression. Multi-Layer (Enhanced) Header Compression 12B 8B Inter-Frame Gap (IFG) Preabmle L1 header (PHY) 4B 4B GFP header Compressed header & Flow ID Payload CRC 6B 6B 2B 2B 2B 2B 2B 24/40B 8/28B 4B MAC DA MAC SA 0x8A88 (opt) S-Vlan (opt) 0x8100 (opt) C-Vlan (opt) 0x0800/0x86DD IPv4/6 UDP/TCP Payload CRC L2 header (MAC) L3 header L4 header MAC IP-10Q s Multi-Layer (enhanced) header compression can improve effective throughput by up to 300% or more without affecting user traffic Enhanced Header Compression Compatibility The IP-10Q s configuration monitoring mechanism is used to provide backwards compatibility with legacy hardware and software versions that do not support Multi-Layer (enhanced) header compression. A configuration mismatch may occur if the remote IDU is configured to Legacy compression mode. In this scenario, both sides of the link will use Legacy compression mode and an alarm will be raised to indicate that there is a configuration mismatch. Ceragon Proprietary and Confidential Page 59 of 225

60 Enhanced Header Compression Counters In order to help operators optimize Multi-Layer (Enhanced) header compression, IP-10Q provides counters when Enhanced Compression is enabled. These counters include real-time information, such as the number of currently active flows and the number of flows by specific flow type. This information can be used by operators to monitor network usage and capacity, and optimize the Multi-Layer compression settings. By monitoring the effectiveness of the compression settings, the operator can adjust these settings to ensure that the network achieves the highest possible effective throughput. Ceragon Proprietary and Confidential Page 60 of 225

61 Ethernet Header Compression Comparison The following table summarizes the basic features of IP-10Q s legacy and enhanced Ethernet header compression mechanisms. Ethernet Header Compression Comparison Table No Compression (L1 header suppression only) MAC (L2) Header Compression (Legacy Mode) Multi-Layer (L2-4) Header Compression (Enhanced Compression) SW license - - Enhanced Compression license required L1 header suppression (removing IFG and Preamble fields) Yes Yes Yes Compressed headers - L2: MAC SA (6 bytes) MAC DA (6 bytes) L2: L3: L4: Ethertype (2 bytes) MAC SA (6 bytes) MAC DA (6 bytes) Outer VLAN header (4 bytes) Inner VLAN header (4 bytes) MPLS header (4 bytes) B-MAC header (22 bytes) IPv4 header (24 bytes) IPv6 header (40 bytes) UDP header (8 bytes) TCP header (28 bytes) Number of flows Ceragon Proprietary and Confidential Page 61 of 225

62 5.3.3 Latency IP-10Q provides best-in-class latency (RFC-2544) for all channels, making it LTE (Long-Term Evolution) ready: <0.2msec for 30/40 MHz channels (1518 byte frames) Benefits of IP-10Q s Top-of-the-Line Low Latency IP-10Q s ability to meet the stringent latency requirements for LTE systems provides the key to expanded broadband wireless services: Longer radio chains Larger radio rings Shorter recovery times More capacity Easing of Broadband Wireless Access (BWA) limitations Frame Cut-Through Support Frames assigned to high priority queues can pre-empt frames already in transmission over the radio from other queues. Transmission of the preempted frames is resumed after the cut-through with no capacity loss or retransmission required. This feature provides services that are sensitive to delay and delay variation, such as VoIP and Pseudowires, with true transparency to lower priority services. Notes: Frame Cut-Through is not supported in the current software release, but is planned for future release. Contact your Ceragon representative for up-to-date information on availability. For additional information: Ethernet Latency Specifications Ceragon Proprietary and Confidential Page 62 of 225

63 5.4 Radio Features This section includes: Adaptive Coding Modulation (ACM) ACM with Adaptive Transmit Power Cross Polarization Interface Canceller (XPIC) Multi-Radio Diversity ATPC Override Timer Ceragon Proprietary and Confidential Page 63 of 225

64 5.4.1 Adaptive Coding Modulation (ACM) This feature cannot be used with the following: BBS Space Diversity 2+0 Multi-Radio with IDU and Line Protection Related topics: ACM with Adaptive Transmit Power Quality of Service (Traffic Manager) Cross Polarization Interface Canceller (XPIC) 1+1 HSB Protection FibeAir IP-10Q employs full-range dynamic ACM. IP-10Q s ACM mechanism copes with 90 db per second fading in order to ensure high transmission quality. IP-10Q s ACM mechanism is designed to work with IP-10Q s QoS mechanism to ensure that high priority voice and data packets are never dropped, thus maintaining even the most stringent service level agreements (SLAs). The hitless and errorless functionality of IP-10Q s ACM has another major advantage in that it ensures that TCP/IP sessions do not time-out. Without ACM, even interruptions as short as 50 milliseconds can lead to timeout of TCP/IP sessions, which are followed by a drastic throughout decrease while these sessions recover. Ceragon Proprietary and Confidential Page 64 of 225

65 Eight Working Points IP-10Q implements ACM with eight available working points, as follows: ACM Working Points (Profiles) Working Point (Profile) Profile 0 Profile 1 Profile 2 Profile 3 Profile 4 Profile 5 Profile 6 Profile 7 Modulation QPSK 8 PSK 16 QAM 32 QAM 64 QAM 128 QAM 256 QAM Strong FEC 256 QAM Light FEC Adaptive Coding and Modulation with Eight Working Points Hitless and Errorless Step-by Step Adjustments ACM works as follows. Assuming a system configured for 128 QAM with ~170 Mbps capacity over a 28 MHz channel, when the receive signal Bit Error Ratio (BER) level reaches a predetermined threshold, the system preemptively switches to 64 QAM and the throughput is stepped down to ~140 Mbps. This is an errorless, virtually instantaneous switch. The system continues to operate at 64 QAM until the fading condition either intensifies or disappears. If the fade intensifies, another switch takes the system down to 32 QAM. If, on the other hand, the weather condition improves, the modulation is switched back to the next higher step (e.g., 128 QAM) and so on, step by step.the switching continues automatically and as quickly as needed, and can reach all the way down to QPSK during extreme conditions. Ceragon Proprietary and Confidential Page 65 of 225

66 Adaptive Coding and Modulation ACM Radio Scripts An ACM radio script is constructed of a set of profiles. Each profile is defined by a modulation order (QAM) and coding rate, and defines the profile s capacity (bps). When an ACM script is activated, the system automatically chooses which profile to use according to the channel fading conditions. The ACM TX profile can be different from the ACM RX profile. The ACM TX profile is determined by remote RX MSE performance. The RX end is the one that initiates an ACM profile upgrade or downgrade. When MSE improves above a predefined threshold, RX generates a request to the remote TX to upgrade its profile. If MSE degrades below a predefined threshold, RX generates a request to the remote TX to downgrade its profile. ACM profiles are decreased or increased in an errorless operation, without affecting Ethernet traffic. ACM scripts can be activated in one of two modes: Fixed Mode. In this mode, the user can select the specific profile from all available profiles in the script. The selected profile is the only profile that will be valid, and the ACM engine will be forced to be OFF. This mode can be chosen without an ACM license. Adaptive Mode. In this mode, the ACM engine is running, which means that the radio adapts its profile according to the channel fading conditions. Adaptive mode requires an ACM license. In the case of XPIC/ACM scripts, all the required conditions for XPIC apply Configurable Maximum and Minimum ACM Profile The user can define both a maximum and a minimum profile. For example, if the user selects a maximum profile of 5, the system will not climb above the profile 5, even if channel fading conditions allow it. If the user selects a minimum profile of 3 (32 QAM), the system will not climb below 32 QAM. If the channel s SNR degrades below the 32 QAM threshold, the radio will lose carrier synchronization, and will report loss of frame. Ceragon Proprietary and Confidential Page 66 of 225

67 ACM Benefits The advantages of IP-10Q s dynamic ACM include: Maximized spectrum usage Increased capacity over a given bandwidth Eight modulation/coding work points (~3 db system gain for each point change) Hitless and errorless modulation/coding changes, based on signal quality Adaptive Radio Tx Power per modulation for maximal system gain per working point An integrated QoS mechanism that enables intelligent congestion management to ensure that high priority traffic is not affected during link fading ACM and Built-In QoS IP-10Q s ACM mechanism is designed to work with IP-10Q s QoS mechanism to ensure that high priority voice and data packets are never dropped, thus maintaining even the most stringent SLAs. Since QoS provides priority support for different classes of service, according to a wide range of criteria, you can configure IP-10Q to discard only low priority packets as conditions deteriorate. If you want to rely on an external switch s QoS, ACM can work with them via the flow control mechanism supported in the radio ACM and 1+1 HSB When ACM is activated together with 1+1 HSB protection, it is essential to feed the active IDU via the main channel of the coupler (lossless channel), and to feed the standby unit via the secondary channel of the coupler (-6db attenuated channel). This maximizes system gain and optimizes ACM behavior for the following reasons: In the TX direction, the power will experience minimal attenuation. In the RX direction, the received signal will be minimally attenuated. Thus, the receiver will be able to lock on a higher ACM profile (according to what is dictated by the RF channel conditions). If the standby IDU is fed via the main channel of the coupler, when the remote unit transmits in QPSK modulation (profile-0), there is a chance that the active unit will have its LOF alarm raised, because its RSL will be 6db below the RSL of the standby unit, while the standby unit will have its LOF alarm cleared. In this scenario, a protection switch is not initiated, even though the active IDU is in LOF, and the standby IDU appears to be functioning normally. When activating an ACM script together with 1+1 HSB protection, if an LOF alarm is raised, both the active and the standby receivers degrade to the lowest available profile (highest RX sensitivity). Because RX sensitivity is very high, the receivers may have false lock, which will result in a switchover. If the LOF alarm remains, protection switchovers may appear alternately every one second. This may cause management instability and may even prevent management access to the units completely. Ceragon Proprietary and Confidential Page 67 of 225

68 In order to avoid this scenario, it is important to carefully follow the instructions for setting up 1+1 HSB protection. In particular, make sure that the link is established with lockout configuration in order to avoid alternate switchovers. Once the link is up and running, lockout can be disabled. The following ACM behavior should be expected in a 1+1 configuration: In the TX direction, the Active TX will follow the remote Active RX ACM requests (according to the remote Active Rx MSE performance). The Standby TX might have the same profile as the Active TX, or might stay at the lowest profile (profile-0). That depends on whether the Standby TX was able to follow the remote RX Active unit s ACM requests (only the active remote RX sends ACM request messages). In the RX direction, both the active and the standby units follow the remote Active TX profile (which is the only active transmitter). Ceragon Proprietary and Confidential Page 68 of 225

69 5.4.2 ACM with Adaptive Transmit Power This feature requires: ACM script ACM enabled prior to enabling ACM with Adaptive Transmit Power RFU-C with software version 2.01 or higher When planning ACM-based radio links, the radio planner attempts to apply the lowest transmit power that will perform satisfactorily at the highest level of modulation. During fade conditions requiring a modulation drop, most radio systems cannot increase transmit power to compensate for the signal degradation, resulting in a deeper reduction in capacity. IP-10Q is capable of adjusting power on the fly, and optimizing the available capacity at every modulation point, as illustrated in the figure below. This figure shows how operators that want to use ACM to benefit from high levels of modulation (e.g., 256 QAM) must settle for low system gain, in this case, 18 db, for all the other modulations as well. With FibeAir IP-10Q, operators can automatically adjust power levels, achieving the extra 4 db system gain that is required to maintain optimal throughput levels under all conditions. The following figure contrasts the transmit output power achieved by using ACM with Adaptive Power to the transmit output power at a fixed power level, over an GHz link. IP-10Q ACM with Adaptive Power Contrasted to Other ACM Implementations For this feature to be used effectively, it is essential for the operator not to breach any regulator-imposed EIRP limitations. For example, if used, the operator must license the system for the maximum possible EIRP. Ceragon Proprietary and Confidential Page 69 of 225

70 5.4.3 Cross Polarization Interface Canceller (XPIC) This feature requires: 2+0 configuration XPIC is one of the best ways to break the barriers of spectral efficiency. Using dual-polarization radio over a single-frequency channel, a dual polarization radio transmits two separate carrier waves over the same frequency, but using alternating polarities. Despite the obvious advantages of dualpolarization, one must also keep in mind that typical antennas cannot completely isolate the two polarizations. In addition, propagation effects such as rain can cause polarization rotation, making cross-polarization interference unavoidable. Dual Polarization The relative level of interference is referred to as cross-polarization discrimination (XPD). While lower spectral efficiency systems (with low SNR requirements such as QPSK) can easily tolerate such interference, higher modulation schemes cannot and require XPIC. IP-10Q s XPIC algorithm enables detection of both streams even under the worst levels of XPD such as 10 db. IP-10Q accomplishes this by adaptively subtracting from each carrier the interfering cross carrier, at the right phase and level. For high-modulation schemes such as 256 QAM, an improvement factor of more than 20 db is required so that cross-interference does not adversely affect performance. In addition, XPIC includes an automatic recovery mechanism that ensures that if one carrier fails, or a false signal is received, the mate carrier will not be affected. This mechanism also ensures that when the failure is cleared, both carriers will be operational XPIC Benefits The advantages of FibeAir IP-10Q s XPIC option include: BER of 10e-6 at a co-channel sensitivity of 5 db Multi-Radio Support Ceragon Proprietary and Confidential Page 70 of 225

71 XPIC Implementation In a single channel application, when an interfering channel is transmitted on the same bandwidth as the desired channel, the interference that results may lead to BER in the desired channel. IP-10Q supports a co-channel sensitivity of 33 db at a BER of 10e-6. When applying XPIC, IP-10Q transmits data using two polarizations: horizontal and vertical. These polarizations, in theory, are orthogonal to each other, as shown in the figure below XPIC - Orthogonal Polarizations In a link installation, there is a separation of 30 db of the antenna between the polarizations, and due to misalignments and/or channel degradation, the polarizations are no longer orthogonal. This is shown in the figure below. XPIC Impact of Misalignments and Channel Degradation Note that on the right side of the figure you can see that CarrierR receives the H+v signal, which is the combination of the desired signal H (horizontal) and the interfering signal V (in lower case, to denote that it is the interfering signal). The same happens in CarrierL = V+h. The XPIC mechanism takes the data from CarrierR and CarrierL and, using a cost function, produces the desired data. Ceragon Proprietary and Confidential Page 71 of 225

72 XPIC Impact of Misalignments and Channel Degradation IP-10Q s XPIC reaches a BER of 10e-6 at a co-channel sensitivity of 5 db! The improvement factor in an XPIC system is defined as the SNR@threshold of 10e-6, with or without the XPIC mechanism Conditions for XPIC XPIC is enabled by loading an XPIC script to the radio in the IDU. In order for XPIC to be operational, all the following conditions must be met: Communications with the RFU are established in both IDUs: An RFU must be connected to each IDU The frequency of both radios should be equal. 1+1 HSB protection must not be enabled. The same script must be loaded in both IDUs. The IDU cannot be in standalone mode. If any of these conditions is not met, an alarm will alert the user. In addition, events will inform the user which conditions are not met XPIC Recovery Mechanism The XPIC mechanism is based on signal cancellation and assumes that both of the transmitted signals are received (with a degree of polarity separation). If for some reason, such as hardware failure, one of the carriers stops receiving a signal, the working carrier may be negatively affected by the received signals, which cannot be canceled in this condition. The purpose of the XPIC recovery mechanism is to save the working link while attempting to recover the faulty polarization. The mechanism works as follows: The indication that the recovery mechanism has been activated is a loss of modem preamble lock, which takes place at SNR~10dB. The first action taken by the recovery mechanism is to cause the remote transmitter of the faulty carrier to mute, thus eliminating the disturbing signal and saving the working link. Ceragon Proprietary and Confidential Page 72 of 225

73 Following this, the mechanism attempts at intervals to recover the failed link. In order to do so, it takes the following actions: The remote transmitter is un-muted for a brief period. The recovery mechanism probes the link to find out if it has recovered. If not, it again mutes the remote transmitter. This action is repeated in exponentially larger intervals. This is meant to quickly bring up both channels in case of a brief channel fade, without seriously affecting the working link if the problem has been caused by a hardware failure. The number of recovery attempts is user-configurable Note: Every such recovery attempt will cause a brief traffic hit in the working link. All the time intervals mentioned above (recovery attempt time, initial time between attempts, multiplication factor for attempt time, number of retries) can be configured by the user, but it is recommended to use the default values. The XPIC recovery mechanism is enabled by default, but can be disabled by the user. Ceragon Proprietary and Confidential Page 73 of 225

74 5.4.4 Multi-Radio This feature requires: 2+0 configuration This feature cannot be used with the following: 1+1 HSB BBS Space Diversity Related topics: 2+0 Multi-Radio and 2+0 Multi-Radio with IDU and Line Protection Automatic State Propagation Multi-Radio enables two separate radio carriers to be shared by a single Ethernet port. This provides an Ethernet link over the radio with double capacity, while still behaving as a single Ethernet interface. The IDUs in a Multi-Radio setup operate in master and slave mode. In Multi-Radio mode, traffic is divided among the two carriers optimally at the radio frame level without requiring Ethernet Link Aggregation, and is not dependent on the number of MAC addresses, the number of traffic flows, or momentary traffic capacity. During fading events which cause ACM modulation changes, each carrier fluctuates independently with hitless switchovers between modulations, increasing capacity over a given bandwidth and maximizing spectrum utilization. The result is 100% utilization of radio resources in which traffic load is balanced based on instantaneous radio capacity per carrier and is independent of data/application characteristics, such as the number of flows or capacity per flow. Typical 2+0 Multi-Radio Link Configuration Ceragon Proprietary and Confidential Page 74 of 225

75 Multi-Radio Basic Operation Multi-radio is available for adjacent pairs of IDUs (slots 1 and 2 or 3 and 4). The left IDU on the floor of the chassis is always the master, and the right IDU on the same floor of the chassis as the master IDU is the slave IDU. In regular 1+0 operation, the radio link of each IDU is represented as Eth5. In Multi-Radio mode, the radio port of the master IDU uses the available bandwidth of both radio channels, while the slave IDU does not have any direct Ethernet connection to its own radio. In other words, the slave IDU does not have an Eth5 interface since the radio resource is being used by the master IDU. The following diagram illustrates the Multi-Radio traffic flow: Slave LVDS LVDS Eth x MODEM MODEM Duplication x Eth & LVDS Master Eth 5 Traffic splitter MODEM MODEM Traffic combiner Eth 5 LVDS LVDS At the transmitting side, outgoing traffic at Eth5 in the master IDU is split between its own radio and that of the slave. Each radio transmits its share of the data. At the receiving side, the slave sends the data it receives to the master, which combines it with the data received from its own radio link, recovering all the data. Data is distributed between the two links at the Layer 1 level in an optimal way. Therefore, the distribution is not dependent on the contents of the Ethernet frames. In addition, the distribution is proportional to the available bandwidth in every link: If both links have the same capacity, half the data will be sent through each link. In ACM conditions, the links could be in different modulations; in this case, data will be distributed proportionally in order to maximize the available bandwidth. Ceragon Proprietary and Confidential Page 75 of 225

76 In order for Multi-Radio to work properly, the two radio links should use the same radio script. Note that in the case of ACM, the links may use different modulations, but the same base script must still be configured in both links Graceful Degradation of Service 2+0 Multi-Radio provides for protection and graceful degradation of service in the event of failure of an RFU or the slave IDU. This ensures that if one link is lost, not all data is lost. Instead, bandwidth is simply reduced until the link returns to service. Graceful degradation in Multi-Radio is achieved by blocking one of the radio links from Multi-Radio data. When a link is blocked, the transmitter does not distribute data to this link and the receiver ignores it when combining. The blocking is implemented independently in each direction, but TX and RX always block a link in a coordinated manner. The following are the criteria for blocking a link: Radio LOF Link ID mismatch Minimum ACM point user configurable (including none) Radio Excessive BER user configurable Radio Signal degrade user configurable User command used to debug a link When a radio link is blocked, an alarm is displayed to users Automatic State Propagation in Multi-Radio Automatic State Propagation (ASP) is used in 1+0 links to quickly close line links in the case of a radio link failure in order to signal the fault to xstp and other protocols. In the case of Multi-Radio, however, the failure of a single link does not necessarily mean that the entire logical link is down. Therefore, the user can configure whether ASP will be initiated upon a single radio failure or only upon a failure of both radios. The line LOS criterion for closing the local line port operates normally in Multi-Radio, since the radio link is not involved. Note that the criterion is applicable for the main unit s line interfaces only. The user-defined ASP parameters can be configured separately for Multi- Radio. Ceragon Proprietary and Confidential Page 76 of 225

77 5.4.5 Diversity This section includes: Diversity Overview Baseband Switching (BBS) Space Diversity IF Combining (IFC) Diversity Type Comparison Ceragon Proprietary and Confidential Page 77 of 225

78 Diversity Overview In long distance wireless links, multipath phenomena are common. Both direct and reflected signals are received, which can cause distortion of the signal resulting in signal fade. The impact of this distortion can vary over time, space, and frequency. This fading phenomenon depends mainly on the link geometry and is more severe at long distance links and over flat surfaces or water. It is also affected by air turbulence and water vapor, and can vary quickly during temperature changes due to rapid changes in the reflections phase. Fading can be flat or dispersive. In flat fading, all frequency components of the signal experience the same magnitude of fading. In dispersive, or frequency selective fading, different frequency components of the signal experience decorrelated fading. Direct and Reflected Signals Space Diversity is a common way to negate the effects of fading caused by multipath phenomena. By placing two separate antennas at a sufficient distance from one another, it is statistically likely that if one antenna suffers from fading caused by signal reflection, the other antenna will continue to receive a viable signal. IP-10Q offers two methods of Space Diversity: Baseband Switching (BBS) Each IDU receives a separate signal from a separate antenna. Each IDU compares each of the received signals, and enables the bitstream coming from the receiver with the best signal. Switchover is errorless ( hitless switching ). IF Combining (IFC) Signals from two separate antennas are combined in phase with each other to maximize the signal to noise ratio. IF Combining is performed in the RFU. Ceragon Proprietary and Confidential Page 78 of 225

79 Diversity Signal Flow Note: Space Diversity configurations offer the option of Ethernet line protection using Multi-Unit LAG. Ceragon Proprietary and Confidential Page 79 of 225

80 Baseband Switching (BBS) Space Diversity This feature requires: Two antennas Two RFUs 1+1 HSB configuration This feature cannot be used with the following: ACM Multi-Radio 2+0 Multi-Radio with IDU and Line Protection BBS Space Diversity requires two antennas and RFUs. The antennas must be separated by approximately 15 to 20 meters. Any RFU type supported by IP- 10Q can be used in a BBS Space Diversity configuration. One RFU sends its signal to the active IDU, while the other RFU sends its signal to the standby IDU. The IDUs share these signals through the chassis backplane, such that each IDU receives data from both RFUs. The diversity mechanism, which is located within the IDU Mux, is active in both IDUs, and selects the better signal based on: Faulty signal indication An indication from the Modem to the Mux, signaling that there are more errors in the traffic stream than it can correct. The purpose of this indication is to alert the Mux to the fact that those errors are on their way, requiring a hitless switchover in order to prevent them from entering the data stream from the Mux onward. OOF (Out-of-Frame) When the Mux identifies an OOF event, it will initiate a switchover. BBS Space Diversity requires a 1+1 configuration in which there are two IDUs and two RFUs protecting each other at both ends of the link. In the event of IDU failure, Space Diversity is lost until recovery, but the system remains protected through the ordinary switchover mechanism IF Combining (IFC) This feature requires: Dual-receiver RFU (FibeAir 1500HP) The RFU receives and processes both signals, and combines them into a single, optimized signal. The IFC mechanism gains up to 2.5 db in system gain. Note: 1500 HP (11 GHz) 40 MHz bandwidth does not support IF Combining. For this frequency, space diversity is only available via BBS. Ceragon Proprietary and Confidential Page 80 of 225

81 Diversity Type Comparison The following table shows the relative benefits and limitations of IFC Space Diversity and BBS Space Diversity. IFC and BBS Space Diversity Comparison IFC BBS Space Diversity RFU Support 1500HP (split mount or all indoor) 7 All Ceragon RFUs Gain Hitless and Errorless Gaining up to 2.5 db in system gain. Hitless and Errorless Does not add to system gain, but is more reliable with sporadic errors. Limitations Symbol rate-dependant. Cannot be used with ACM or Multi-Radio. Configurations N+0 N HP (11 GHz ) 40 MHz bandwidth does not support IF Combining. For this frequency, space diversity is only available via BBS. Ceragon Proprietary and Confidential Page 81 of 225

82 5.4.6 ATPC Override Timer ATPC is a closed-loop mechanism by which each RFU changes the transmitted signal power according to the indication received across the link, in order to achieve a desired RSL on the other side of the link. Without ATPC, if loss of frame occurs the system automatically increases its transmit power to the configured maximum. This may cause a higher level of interference with other systems until the failure is corrected. In order to minimize this interference, some regulators require a timer mechanism which will be manually overridden when the failure is fixed. The underlying principle is that the system should start a timer from the moment maximum power has been reached. If the timer expires, ATPC is overridden and the system transmits at a pre-determined power level until the user manually re-establishes ATPC and the system works normally again. The user can configure the following parameters: Override timeout (0 to disable the feature): The amount of time the timer counts from the moment the system transmits at the maximum configured power. Override transmission power: The power that will be transmitted if ATPC is overridden because of timeout. The user can also display the current countdown value. When the system enters into the override state, ATPC is automatically disabled and the system transmits at the pre-determined override power. An alarm is raised in this situation. The only way to go back to normal operation is to manually cancel the override. When doing so, users should be sure that the problem has been corrected; otherwise, ATPC may be overridden again. Ceragon Proprietary and Confidential Page 82 of 225

83 5.5 Ethernet Features This section includes: Automatic State Propagation Ceragon Proprietary and Confidential Page 83 of 225

84 5.5.1 Automatic State Propagation Related topics: Multi-Radio Automatic State Propagation ("GigE Tx mute override") enables propagation of radio failures back to the line, to improve the recovery performance of resiliency protocols (such as xstp). The feature enables the user to configure which criteria will force the GbE port (or ports in case of a remote fault) to be muted or shutdown, in order to allow the network to find alternative paths. Upon radio failure, Eth1 is muted when configured as optical or shut down when configured as electrical. In 2+0 Multi-Radio mode, Automatic State Propagation can be triggered upon a failure in a single IDU or upon a failure in both IDUs. This behavior is determined by user configuration. Automatic State Propagation Port Behavior User Configuration Optical (SFP) GbE Port Behavior Electrical GbE port (10/100/1000) Port Behavior Automatic State Propagation disabled. Local LOF, Link-ID mismatch (always enabled) Ethernet shutdown threshold profile. Local Excessive BER Local LOC No mute is issued. Mute the LOCAL port when one or more of the following events occurs: 1. Radio-LOF on the LOCAL unit. 2. Link ID mismatch on the LOCAL unit. Mute the LOCAL port when ACM Rx profile degrades below a pre-configured profile on the LOCAL unit Mute the LOCAL port when an Excessive BER alarm is raised on the LOCAL unit Mute the LOCAL port when a GbE-LOC alarm is raised on the LOCAL unit. No shutdown. Shut down the LOCAL port when one or more of the following events occurs: 1. Radio-LOF on the LOCAL unit. 2. Link ID mismatch on the LOCAL unit. Shut down the LOCAL port when ACM Rx profile degrades below a pre-configured profile on the LOCAL unit. This capability is applicable only when ACM is enabled. Shut down the LOCAL port when an Excessive BER alarm is raised on the LOCAL unit No shutdown. Note1: Electrical-GbE cannot be muted. Electrical-GbE LOC will not trigger Shutdown, because it will not be possible to enable the port when the LOC alarm is cleared Ceragon Proprietary and Confidential Page 84 of 225

85 User Configuration Optical (SFP) GbE Port Behavior Electrical GbE port (10/100/1000) Port Behavior Remote Fault Mute the LOCAL port when one or more of the following events is raised on the REMOTE unit: 1. Radio-LOF (on remote). 2. Link-ID mismatch (on remote). 3. GbE-LOC alarm is raised (on remote). 4. ACM Rx profile crossing threshold (on remote), only if enabled on the LOCAL. 5. Excessive BER (on remote), only if enabled on the LOCAL. Shut down the LOCAL port, when one or more of the following events is raised on the REMOTE unit: 1. Radio-LOF (on remote). 2. Link-ID mismatch (on remote). 3. ACM Rx profile crossing threshold (on remote), only if enabled on the LOCAL. 4. Excessive BER (on remote), only if enabled on the LOCAL. Note1: Electrical-GbE cannot be muted. Electrical-GbE LOC will not trigger "Shutdown", because it will not be possible to enable the port when LOC alarm is cleared Notes: It is recommended to configure both ends of the link to the same Automatic State Propagation configuration. If the link uses In-Band management, when the port is muted or shut down, management distributed through the link might be lost. If this occurs, the unit will not be manageable. The unit will only become manageable again when the port is un-muted or enabled. Ceragon Proprietary and Confidential Page 85 of 225

86 5.6 Quality of Service (Traffic Manager) This section includes: Integrated Quality of Service (QoS) Overview Standard QoS Enhanced QoS Standard and Enhanced QoS Comparison Ceragon Proprietary and Confidential Page 86 of 225

87 5.6.1 Integrated Quality of Service (QoS) Overview Related topics: Standard and Enhanced QoS Comparison IP-10Q offers integrated QoS functionality. In addition to its standard QoS functionality, IP-10Q offers an enhanced QoS feature. Enhanced QoS is licenseactivated. IP-10Q s standard QoS provides for four queues and six classification criteria. Ingress traffic is limited per port, Class of Service (CoS), and traffic type. Scheduling is performed according to Strict Priority (SP), Weighted Round Robin (WRR), or Hybrid WRR/SP scheduling. IP-10Q s enhanced QoS provides eight classification criteria instead of six, color-awareness, increased frame buffer memory, eight priority queues with configurable buffer length, improved congestion management using WRED protocols, enhanced counters, and other enhanced functionality. The figure below shows the QoS flow of traffic. QoS Traffic Flow Ceragon Proprietary and Confidential Page 87 of 225

88 5.6.2 Standard QoS QoS enables users to configure classification and scheduling to ensure that packets are forwarded and discarded according to their priority. Since it is common to set QoS and rate limiting settings identically in several ports, the QoS configuration can be copied from one port to another. This saves considerable time and prevents configuration mistakes. The following diagram illustrates the QoS flow: Ingress Port #x Egress Port #y Classifier (4 Queues) 5 Policers (Ingress Rate Limiting) Marker Queue Controller Scheduler Shaper (Egress rate limiting) Standard QoS Classifier Using IP-10Q s standard QoS functionality, the system examines the incoming traffic and assigns the desired priority according to the marking of the packets (based on the user port/l2/l3 marking in the packet). In case of congestion in the ingress port, low priority packets are discarded first. The standard QoS classifier is made up of four classification criteria hierarchies: MAC DA (Destination Address) Overwrite Classification and marking is performed for incoming frames carrying a MAC DA that appears in the Static MAC table, according to the following options: Disable No MAC DA classification or VLAN P-Bit overwrite (marking). Queue Decision Only classification to queue. No marking. VLAN P-Bit Overwrite Only VLAN P-Bits overwrite (marking). Classification according to a lower criterion. Queue Decision and VLAN P-Bit Overwrite Both classification and VLAN P-Bits overwrite. VLAN ID Overwrite If the first criteria is not fulfilled (either because it is disabled, or because the ingress frame does not carry any MAC DA that appears in the S MAC table), classification and/or marking (VLAN P-Bit overwrite, assuming the frame egress is tagged) is decided according to the VLAN ID to Queue table according to the following options: Disable No VLAN ID classification or VLAN P-Bit overwrite (marking). Queue Decision Only classification to queue. No marking. VLAN P-Bit Overwrite Only VLAN P-Bit overwrite (marking). Classification is according to the lower criteria (P-Bits or port priority). In this case, P-Bits are assigned as follows (if egress frame is tagged): Frames classified to 1 st queue are given p-bits=0 Frames classified to 2 nd queue are given p-bits=2 Frames classified to 3 rd queue are given p-bits=4 Frames classified to 4 th queue are given p-bits=6 Ceragon Proprietary and Confidential Page 88 of 225

89 Queue Decision and VLAN P-Bit Overwrite Both classification and VLAN P-Bit overwrite. Initial Classification is according to the following configuration: VLAN P-Bit Classification is according to VLAN P-Bit. And the queue is assigned according to the VLAN P-Bit to Queue table. IP TOS Classification is according to IP TOS (IP precedence, or IP diffserv). The queue is assigned according to the IP P-Bit to Queue table. VLAN P-Bit over IP TOS Classification according to VLAN P-Bit, if the ingress frame carries a VLAN. For untagged packets with an IP header, classification is according to IP TOS. IP TOS over VLAN P-Bit Classification is according to IP TOS, if the ingress frame has an IP header. If the ingress frame without an IP header carries a VLAN, classification is according to VLAN P-Bit. Port (Default) If any of the above criteria are not fulfilled, the default classification is assigned to the ingress frame according to the port priority. Default Classification. Default priority for frames incoming at the port Standard QoS Policers IP-10Q s standard QoS provides up to five policers to perform ingress rate limiting. The policers are based on a color blind leaky bucket scheme, and can be applied per port or CoS. For each policer, users can define up to five class maps. Each class map includes the following parameters: Committed Information Rate (CIR) IP-10Q supports CIR granularity of 64kbps up to 1 Mbps of CIR, 1 Mbps from 1 Mbps to 1 Gbps of CIR. Packets within the CIR defined for the service are marked Green and passed through the QoS module. Committed Burst Size (CBS) IP-10Q supports CBS up to a maximum of 128 kbytes. The default value is 12 kbytes. Packets within the CBS defined for the service are marked Green and passed through the QoS module. Committed Information Rate (CIR) IP-10Q supports the following granularity for CIR: 64Kbps <= CIR <= 960Kbps, in steps of 64Kbps. 1000Kbps <= CIR <= 100,000Kbps in steps of 1000Kbps. 100,000Kbps < CIR <= 1,000,000Kbps in steps of 10,000Kbps. Committed Burst Size (CBS) IP-10Q supports the following granularity for CBS: For 64Kbps <= CIR <= 960Kbps, 0 < CBS <= 273,404 Bytes. For 1000Kbps <= CIR <= 100,000Kbps, 0 < CBS <= 132,585 Bytes. For 100,000Kbps < CIR <= 1,000,000Kbps, 0 < CBS <= 4,192,668 Bytes. Ceragon Proprietary and Confidential Page 89 of 225

90 Data type The rate can be limited based on the following data types: None (no limiting), Unknown unicast, Unknown multicast, Broadcast, Multicast, Unicast, Management, ARP, TCP-Data, TCP-Control, UDP, Non- UDP, Non-TCP-UDP, Queue1, Queue2, Queue3, Queue4. Note: Note: Limit Exceed Action Discard Frame. Management frames are BPDUs processed by the system s IDC, when processing L2 protocols (e.g., xstp). The rate for rate limiting is measured for all Layer 1 bytes, meaning: Preamble (8bytes) + Frame's DA to CRC + IFG (12 Bytes) Queue Management, Scheduling, and Shaping IP-10Q s standard QoS has four priority queues. The queue controller distributes frames to the queues according to the classifier. The fourth queue is the highest priority queue, and the first queue is the lowest priority queue. The scheduler determines how frames are output from the queues. IP-10Q s standard QoS supports the following scheduling schemes: Strict Priority for all queues. Strict Priority for the fourth queue, and Weighted Round Robin (WRR) for the remaining queues. Strict Priority for the fourth and third queues, and WRR for second and first queues. WRR for all queues. In a WRR scheduling scheme, a weight is assigned to each queue, so that frames egress from the queues according to their assigned weight, in order to avoid starvation of lower priority queues. In addition, frames egress in a mixed manner, in order to avoid bursts of frames from the same queue. Each queue s weight can be configured. A queue's weight is used by the scheduler when the specific queue is part of a WRR scheduling scheme. Queue-Weight can be configured in the range of The default queue weights are 8,4,2,1. The shaper determines the scheduler rate (egress rate limit). The shaper can be enabled and disabled by the user. By default, the shaper is disabled. The shaper rate is set with the following granularity: For 64Kbps <= Rate <= 960Kbps, in steps of 64Kbps. For 1000Kbps <= Rate <= 100,000Kbps in steps of 1000Kbps. For 100,000Kbps < Rate <= 1,000,000Kbps in steps of 10,000Kbps. Ceragon Proprietary and Confidential Page 90 of 225

91 5.6.3 Enhanced QoS This feature requires: Enhanced QoS license Related topics: Synchronization Using Precision Timing Protocol (PTP) Optimized Transport Licensing Enhanced QoS provides an enhanced and expanded feature set. The tools provided by enhanced QoS apply to egress traffic on the radio port, which is where bottlenecks generally occur. Enhanced QoS can be enabled and disabled by the user. Enhanced QoS capabilities include: Enhanced classification criteria Eight priority queues with configurable buffer length An enhanced scheduler based on Strict Priority, Weighted Fair Queue (WFQ), or a hybrid approach that combines Strict Priority and WFQ Shaper per priority queue WRED support, along with Tail-Drop, for congestion management Configurable P-bit and CFI/DEI re-marker A PTP Optimized Transport dedicated channel for time synchronization protocols Enhanced counters These and other IP-10Q enhanced QoS features enable operators to provide differentiated services with strict SLA while maximizing network resource utilization. Enhanced QoS requires a license, and can be enabled and disabled by the user. The main benefits of enhanced QoS are: Improved available link capacity utilization: Enhanced and configurable queue buffer size (4 Mb total) WRED for best utilization of the link when TCP/IP sessions are transported, providing up to 25% more capacity. Enhanced service differentiation: 8 CoS queues (as opposed to 4 queues in standard QoS) Additional classification criteria MPLS EXP bits and UDP ports Shaping per CoS queue Sync. Optimized transport - best performance for 1588 packets Monitoring, Assurance and Diagnostics capabilities: Per queue counters Transmitted and dropped traffic Ceragon Proprietary and Confidential Page 91 of 225

92 The following figure illustrates the basic building blocks and traffic flow of enhanced QoS. IP-10Q Enhanced QoS The initial step in the enhanced QoS traffic flow is the classifier, which provides granular service classification based on a number of user-defined criteria. The classifier marks the Service ID, CoS, and color of the frames. If a frame s VLAN ID matches a Service ID that is mapped to a policer, the frame is sent to the policer. Untagged frames or frames whose VLAN ID does not match a defined Service ID are sent directly to a queue, based on the frame s CoS and color. The next step is queue management. Queue management determines which packets enter which of the eight available queues. Queue management also includes congestion management, which can be implemented by Tail-Drop or WRED. Frames are sent out of the queues according to scheduling and shaping, IP- 10Q s enhanced QoS module provides a unique hierarchical scheduling model that includes four priorities, with WFQ within each priority and shaping per queue. This model enables operators to define flexible and highly granular QoS schemes for any mix of services. Finally, the enhanced QoS module re-marks the P-bits and CFI/DEI bits of the most outer VLAN according to the CoS and color decision in the classifier. This step is also known as the modifier. Ceragon Proprietary and Confidential Page 92 of 225

93 Enhanced QoS Classifier The classifier is a basic element of each QoS mechanism. Each frame is assigned a Class of Service (CoS) and color, based on MEF 10.2 recommendations. The user can define several criteria by which frames are classified. Classifier Traffic Flow Each frame is assigned a CoS and Color CoS is a 3-bit value from 0-7 that is used for classification to priority queues. Color is a 1-bit value (Green or Yellow) used for policing. Green represents CIR, and Yellow represents EIR. Classification to CoS and Color can be based on the following criteria Note: First hierarchy Based on destination MAC address or source/destination UDP ports. The first classification hierarchy is used to identify and give priority to network protocols. Layer2 protocols such as xstp and Slow protocols can be classified based on their pre-defined destination MAC address. Higher layer protocols such as NTP can be identified based on UDP ports. Second hierarchy Based on VLAN ID. The second hierarchy is used to classify frames based on network services. Each service is assigned to a different VLAN. Frames can be also prioritized based on their in-band management VLAN ID. To prevent loss of management to the remote sites, classification by In-Band management must be configured before activating the enhanced QoS feature. Especially at the first activation after upgrade, the In-Band management VLAN ID should be assigned CoS 7 and Green color. Third hierarchy Based on Priority bits. Options are VLAN 802.1p p-bits, IP DSCP/TOS, and/or MPLS experimental bits. Classification is performed in the order of cardinality listed above. The classifier checks the first hierarchy, the second hierarchy, and the third hierarchy, until a match is found. Ceragon Proprietary and Confidential Page 93 of 225

94 Each frame is assigned a Service ID Classification to Services is based on VLAN ID. A Service ID is used for policing and for classification to CoS. Each policer is monitored by statistics counters. Each CoS is mapped to one of the 8 available priority queues All the classification criteria are divided into three hierarchies according to their cardinality, from the most specific to the most general. Each queue is assigned a priority Priorities vary from the highest (fourth) to the lowest (first). The scheduling mechanism treats these priorities as strict. WFQ scheduling is performed between queues of the same priority. For detailed information about scheduling, refer to Scheduling and Shaping on page Queue Management Queue management is the process by which packets are assigned to priority queues. Queue management also includes congestion management. IP-10Q provides the tail-drop method of congestion management, and enhanced QoS also offers Weighted Random Early Detection (WRED). Enhanced QoS supports eight queues with configurable buffer size. The user can specify the buffer size of each queue independently. The total amount of memory dedicated to these queue buffers is 4Mb, and the size of each queue can be set to 0.5, 1, 2, or 4Mb. The default buffer size is 0.5Mb for each queue. The following considerations should be taken into account in determining the proper buffer size: Note: Latency considerations If low latency is required (users would rather drop frames in the queue than increase latency) small buffer sizes are preferable. The actual, effective buffer size of the queue can be less than 0.5Mb based on the configuration of the WRED tail drop curve. Throughput immunity to fast bursts When traffic is characterized by fast bursts, it is recommended to increase the buffer sizes of the priority queues to prevent packet loss. Of course, this comes at the cost of a possible increase in latency. User can configure burst size as a tradeoff between latency and immunity to bursts, according the application requirements. One of the key features of IP-10Q s enhanced QoS is the use of WRED to manage congestion scenarios. WRED provides several advantages over the standard tail-drop congestion management method. WRED enables differentiation between higher and lower priority traffic based on CoS. Moreover, WRED can increase capacity utilization by eliminating the phenomenon of global synchronization. Global synchronization occurs when TCP flows sharing bottleneck conditions receive loss indications at around the same time. This can result in periods during which link bandwidth utilization drops significantly as a consequence of a simultaneous falling to a slow start Ceragon Proprietary and Confidential Page 94 of 225

95 of all the TCP flows. The following figure demonstrates the behavior of two TCP flows over time without WRED. Synchronized Packet Loss WRED eliminates the occurrence of traffic congestion peaks by restraining the transmission rate of the TCP flows. Each queue occupancy level is monitored by the WRED mechanism and randomly selected frames are dropped before the queue becomes overcrowded. Each TCP flow recognizes a frame loss and restrains its transmission rate (basically by reducing the window size). Since the frames are dropped randomly, statistically each time another flow has to restrain its transmission rate as a result of frame loss (before the real congestion occurs). In this way, the overall aggregated load on the radio link remains stable while the transmission rate of each individual flow continues to fluctuate similarly. The following figure demonstrates the transmission rate of two TCP flows and the aggregated load over time when WRED is enabled. Random Packet Loss with Increased Capacity Utilization Using WRED Each one of the eight priority queues can be given a different weight. For each queue, the user defines the WRED profile curve. This curve describes the probability of randomly dropping frames as a function of queue occupancy. Basically, as the queue occupancy grows, the probability of dropping each incoming frame increases as well. As a consequence, statistically more TCP flows will be restrained before traffic congestion occurs. Ceragon Proprietary and Confidential Page 95 of 225

96 The WRED profile curve can be adjusted for each one of the priority queues. Yellow and Green frames can also be assigned different weights. Usually, Green frames (committed rate) are preferred over Yellow frames (excessive rate), as shown in the curve below. WRED Profile Curve Note: WRED can also be set to a tail drop curve. A tail drop curve is useful for reducing the effective queue size, such as when low latency must be guaranteed. In order to set the tail drop curve to its maximum level, the drop percentage must be set to zero Scheduling and Shaping Scheduling and shaping determine how traffic is sent on to the radio from the queues. Scheduling determines the priority among the queues, and shaping determines the traffic profile for each queue. IP-10Q s enhanced QoS module provides a unique hierarchical scheduling model that includes four priorities, with Weighted Fair Queuing (WFQ) within each priority, and shaping per port and per queue. This model enables operators to define flexible and highly granular QoS schemes for any mix of services. Shaping The egress shaper is used to shape the traffic profile sent to the radio. In enhanced QoS mode, there is an egress shaper for each priority queue. The user can configure CIR, CBS, and line compensation. Note: The user can configure the shaper to count in L2 by setting line compensation to zero. The user can also punish short frame senders for the overhead they cause in the network by increasing the line compensation to a value above 20 bytes. Ceragon Proprietary and Confidential Page 96 of 225

97 Scheduling IP-10Q s enhanced QoS mechanism provides Strict Priority and Weighted Fair Queue (WFQ) for scheduling. Users can configure a combination of both methods to achieve the optimal results for their unique network requirements. Each priority queue has a configurable strict priority from 1 to 4 (4=High;1=Low). WFQ weights are used to partition bandwidth between queues of the same priority. Queue Priority Configuration Example For each queue, the user configures the following parameters: Note: Priority (1 to 4) The priority value is strictly applied. This means the queue with higher priority will egress before a queue with lower priority, regardless of WFQ weights. WFQ weight (1 to 15) Defines the ratio between the bandwidth given to queues of the same priority. For example if queue 6 and queue 7 are assigned WFQ weights of 4 and 8, respectively (using the notations of the above figure), then under congestion conditions queue 7 will be allowed to transmit twice as much bandwidth as queue 6. In order to be able to egress frames, each queue must also have enough credits in its shaper. Ceragon Proprietary and Confidential Page 97 of 225

98 Scheduling Examples This section provides several use cases in which Strict Priority and WFQ are combined to produce a desired scheduling configuration. These are simply two examples of the many ways in which IP-10Q s flexible scheduling mechanism can be configured to achieve a combination of Strict Priority scheduling for the highest priority traffic flows and weighted scheduling for other traffic flows that may be less delay sensitive. Example 1 shows a hybrid setup in which the three highest-priority queues are served according to Strict Priority, and the remaining queues are served according to WFQ. In this example, higher-priority queues are served first. Only after the three highest-priority queues are empty is traffic from the remaining five queues served, according to WFQ and their respective weight. Example 1 Hybrid Scheduling Queue Priority Weight Priority Scheme Strict Priority served according to priority (descending) WFQ - Same priority served according to weight (16 bytes of Q4, 8 bytes of Q5, 4 bytes of Q6, etc.) Example 1 Hybrid Scheduling Illustration Ceragon Proprietary and Confidential Page 98 of 225

99 Example 2 shows a hierarchical scheme in which the highest priority queue is served first, and other queues are only served after the highest-priority queue is empty, according to their respective priorities and weights. Example 2 Hierarchical Scheduling Queue Priority Weight Priority Scheme Highest priority served first Same priority, same weight, evenly serving 1 byte of Q2 and 1 byte of Q Same priority, different weight, serving bytes of Q4 and 1 byte of Q Same priority, different weight, serving bytes of Q6, 2 bytes of Q7 and 1 byte of Q Example 1 Hierarchical Scheduling Illustration Configurable P-Bit and CFI/DEI Re-Marking When enabled, the re-marker modifies each packet s 802.1p P-Bit and CFI/DEI bit fields p is modified according to the classifier decision. The CFI/DEI (color) field is modified according to the classifier and policer decision. The color is first determined by a classifier and may be later overwritten by a policer. Green color is represented by a CFI/DEI value of 0, and Yellow color is represented by a CFI/DEI value of 1. Ceragon Proprietary and Confidential Page 99 of 225

100 5.6.4 Standard and Enhanced QoS Comparison The following table summarizes the basic features of IP-10Q s standard and enhanced QoS functionality. IP-10Q Standard and Enhanced QoS Features Feature Standard QoS Enhanced QoS License Required No Yes Number of CoS Queues 4 8 (radio only) Frame Buffer Size 1 MBit 4 Mbit (on egress port towards radio only), and configurable CoS Classification Criteria Source Port VLAN 802.1p MAC DA IPv4 DSCP/TOS IPv6 TC Additional classification criteria: UDP Port MPLS EXP bits Scheduling Method Strict Priority, Weighted Round Robin (WRR), or Hybrid Four scheduling priorities with WFQ between queues in the same priority Shaping Per port Per queue Congestion Management Tail-drop Tail-drop, and Weighted Random Early Discard (WRED) CIR/EIR Support (Color- Awareness) CIR only CIR + EIR (WRED) CoS to P-bit Re-Marking Default mapping only Color-aware PMs and Statistics RMON Statistics Number of bytes accepted and number of packets dropped. Ceragon Proprietary and Confidential Page 100 of 225

101 5.7 Synchronization This section includes: Synchronization Overview IP-10Q Synchronization Solution Synchronization Using Precision Timing Protocol (PTP) Optimized Transport SyncE PRC Pipe Regenerator Mode Ceragon Proprietary and Confidential Page 101 of 225

102 5.7.1 Synchronization Overview Synchronization is an essential part of any mobile backhaul solution and is sometimes required by other applications as well. Two unique synchronization issues must be addressed for mobile networks: Frequency Lock: Applicable to GSM and UMTS-FDD networks. Limits channel interference between carrier frequency bands. Typical performance target: frequency accuracy of < 50 ppb. Sync is the traditional technique used, with traceability to a PRS master clock carried over PDH/SDH networks, or using GPS. Phase Lock with Latency Correction: Applicable to CDMA, CDMA-2000, UMTS-TDD, and WiMAX networks. Limits coding time division overlap. Typical performance target: frequency accuracy of < ppb, phase difference of < 1-3 ms. GPS is the traditional technique used Precision Timing-Protocol (PTP) PTP synchronization refers to the distribution of frequency, phase, and absolute time information across an asynchronous packet switched network. PTP can use a variety of protocols to achieve timing distribution, including: IEEE-1588 NTP RTP Precision Timing Protocol (PTP) Synchronization Ceragon Proprietary and Confidential Page 102 of 225

103 Synchronous Ethernet (SyncE) SyncE is standardized in ITU-T G.8261 and refers to a method whereby the clock is delivered on the physical layer. The method is based on SDH/TDM timing, with similar performance, and does not change the basic Ethernet standards. The SyncE technique supports synchronized Ethernet outputs as the timing source to an all-ip BTS/NodeB. This method offers the same synchronization quality provided over DS1 interfaces to legacy BTS/NodeB. Synchronous Ethernet (SyncE) IP-10Q Synchronization Solution Ceragon's synchronization solution ensures maximum flexibility by enabling the operator to select any combination of techniques suitable for the operator s network and migration strategy. PTP optimized transport: Supports a variety of protocols, such as IEEE-1588 and NTP Guaranteed ultra-low PDV (<0.035 ms per hop) Unique support for ACM and narrow channels SyncE Regenerator mode PRC grade (G.811) performance for pipe ( regenerator ) applications Ceragon Proprietary and Confidential Page 103 of 225

104 5.7.3 Synchronization Using Precision Timing Protocol (PTP) Optimized Transport This feature requires: Enhanced QoS license Related topics: Enhanced QoS IP-10Q supports the PTP synchronization protocol (IEEE-1588). IP-10Q s PTP Optimized Transport guarantees ultra-low PDV (<0.035 ms), and provides unique support for ACM and narrow channels. Frame delay variation of <0.035 ms per hop for PTP control frames is supported, even when ACM is enabled, and even when operating with narrow radio channels. The Precision Time Protocol (PTP) optimized transport feature is essential for timing synchronization protocols such as IEEE The PTP optimized transport channel is a Constant Bit Rate Channel that is dedicated to the Precision Time protocol with a constant latency that is unaffected by ACM profile changes and by congestion conditions that may occur on the payload traffic path. Ceragon's unique PTP Optimized Transport mechanism ensures that PTP control frames (IEEE-1588, NTP, etc.) are transported with maximum reliability and minimum delay variation, to provide the best possible timing accuracy (frequency and phase) meeting the stringent requirement of emerging 4G technologies. PTP control frames are identified using the advanced integrated QoS classifier. Upon enabling this feature, a special low PDV channel is created. This channel has 2 Mb bandwidth and carries all the frames mapped to the eighth Enhanced QoS priority queue. Once enabling the feature, the user must make sure to classify all PTP frames to the eighth queue. In this mode, all frames from the eight queue will bypass the shaper and scheduler and will be sent directly to the dedicated low PDV channel. Ceragon Proprietary and Confidential Page 104 of 225

105 5.7.4 SyncE PRC Pipe Regenerator Mode Related topics: Licensing In SyncE PRC pipe regenerator mode, frequency is transported between the GbE interfaces through the radio link. PRC pipe regenerator mode makes use of the fact that the system is acting as a simple link (so no distribution mechanism is necessary) in order to achieve the following: Improved frequency distribution performance: PRC quality No use of bandwidth for frequency distribution Simplified configuration For this application IP-10Q has a dedicated mechanism which provides PRC grade (G.811) performance. Ceragon Proprietary and Confidential Page 105 of 225

106 6. Radio Frequency Units (RFUs) This chapter includes: RFU Overview RFU Selection Guide RFU-C 1500HP/RFU-HP RFH-HS RFU-SP 1500P Ceragon Proprietary and Confidential Page 106 of 225

107 6.1 RFU Overview FibeAir Radio Frequency Units (RFUs) were designed with sturdiness, power, simplicity, and compatibility in mind. These advanced systems provide highpower transmission for short and long distances and can be assembled and installed quickly and easily. Any of the RFUs described in this chapter can be used in an IP-10Q system. FibeAir RFUs deliver the maximum capacity over MHz channels with configurable modulation schemes from QPSK to 256QAM. The RFU supports low to high capacities for traditional voice, mission critical, and emerging Ethernet services, with any mix of interfaces, pure Ethernet, pure TDM, or hybrid Ethernet and TDM interfaces (Native 2 ). High spectral efficiency can be ensured with XPIC, using the same bandwidth for double the capacity, via a single carrier, with vertical and horizontal polarizations. IP-10Q works with the following RFUs: Standard Power FibeAir RFU-C FibeAir RFU-SP FibeAir 1500P High Power FibeAir 1500HP FibeAir RFU-HP FibeAir RFU-HS The following RFUs can be installed in a split-mount configuration: FibeAir RFU-C (6 38 GHz) 8 FibeAir 1500HP/RFU-HP (6 11 GHz) FibeAir RFU-HS (6 8 GHz) FibeAir RFU-SP (6 8 GHz) FibeAir 1500P (11 38 GHz) The following RFUs can be installed in an all-indoor configuration: FibeAir 1500HP/RFU-HP (6 11 GHz) The IFU and RFU are connected by a coaxial cable RG-223 (100 m/300 ft), Belden 9914/RG-8 (300 m/1000 ft) or equivalent, N-type connectors (male). The antenna connection can be: Direct or remote mount using the same antenna type. Remote mount: standard flexible waveguide (frequency dependent) 8 Refer to RFU-C roll-out plan for availability of each frequency. Ceragon Proprietary and Confidential Page 107 of 225

108 6.2 RFU Selection Guide The following table can be used to help you select the RFU that is appropriate to your location. For the GHz frequency range, use FibeAir RFU-C For the low frequencies please refer to the options below: RFU Selection Guide Character RFU-C (6 38GHz) 1500HP (6 11GHz) RFU-HP (6 8GHz) RFU-HS (6 8GHz) RFU-SP (6 8GHz) 1500P (11 38GHz) Installation Type Space Diversity Method Frequency Diversity Split Mount All-Indoor SD (BBS/IFC) BBS BBS + IFC 9 BBS BBS BBS BBS FD (BBS) 1+0/2+0/1+1/2+2 Configuration N N+0 ( N>2) Tx Power (dbm) High Power (up to 29 dbm) Ultra High Power (up to 32 dbm) RFU Mounting Direct Mount Antenna Bandwidth (BW) 3.5MHz 56 MHz MHz 30 MHz 56 MHz -- Power Saving Mode Adjustable Power Consumption HP (11 GHz ) 40 MHz bandwidth does not support IF Combining. For this frequency, Space Diversity is only available via BBS. Ceragon Proprietary and Confidential Page 108 of 225

109 6.3 RFU-C FibeAir RFU-C is a fully software configurable, state-of-the-art RFU that supports a broad range of interfaces and capacities from 10 Mbps up to 500 Mbps. RFU-C operates in the frequency range of 6-38 GHz. RFU-C supports low to high capacities for traditional voice and Ethernet services, as well as PDH/SDH/SONET or hybrid Ethernet and TDM interfaces. Traffic capacity throughput and spectral efficiency are optimized with the desired channel bandwidth. For maximum user choice flexibility, channel bandwidths can be selected together with a range of modulations from QPSK to 256 QAM. With RFU-C, traffic capacity throughput and spectral efficiency are optimized with the desired channel bandwidth. For maximum user choice flexibility, channel bandwidths can be selected together with a range of modulations from QPSK to 256 QAM over MHz channel bandwidth. When RFU-C operates in co-channel dual polarization (CCDP) mode using XPIC, two carrier signals can be transmitted over a single channel, using vertical and horizontal polarization. This enables double capacity in the same spectrum bandwidth Main Features of RFU-C Frequency range Operates in the frequency range 6 38 GHz Frequency accuracy ±4 ppm 10 More power in a smaller package - Up to 24 dbm for extended distance, enhanced availability, use of smaller antennas Configurable Modulation QPSK 256 QAM Configurable Channel Bandwidth 3.5 MHz 56MHz Compact, lightweight form factor - Reduces installation and warehousing costs Supported configurations: 1+0 direct and remote mount 1+1 direct and remote mount 2+0 direct and remote mount 2+2 remote mount 4+0 remote mount Efficient and easy installation - Direct mount installation with different antenna types 10 Over temperature. Ceragon Proprietary and Confidential Page 109 of 225

110 6.3.2 RFU-C Frequency Bands Frequency Band TX Range RX Range Tx/Rx Spacing A A A 6L GHz 6H GHz B 252A 240A C 340B Ceragon Proprietary and Confidential Page 110 of 225

111 Frequency Band TX Range RX Range Tx/Rx Spacing A A A 7 GHz C 168B A Ceragon Proprietary and Confidential Page 111 of 225

112 Frequency Band TX Range RX Range Tx/Rx Spacing P O M K J I Ceragon Proprietary and Confidential Page 112 of 225

113 Frequency Band TX Range RX Range Tx/Rx Spacing F D C B Ceragon Proprietary and Confidential Page 113 of 225

114 Frequency Band TX Range RX Range Tx/Rx Spacing A C B A Ceragon Proprietary and Confidential Page 114 of 225

115 Frequency Band TX Range RX Range Tx/Rx Spacing A A 500A 311C-J B 8 GHz A 310D 310C Ceragon Proprietary and Confidential Page 115 of 225

116 Frequency Band TX Range RX Range Tx/Rx Spacing A A C 266B A A A A 10 GHz A 350B 550A Ceragon Proprietary and Confidential Page 116 of 225

117 Frequency Band TX Range RX Range Tx/Rx Spacing A GHz All A GHz GHz Ceragon Proprietary and Confidential Page 117 of 225

118 Frequency Band TX Range RX Range Tx/Rx Spacing GHz GHz / UL GHz All 26 GHz Ceragon Proprietary and Confidential Page 118 of 225

119 Frequency Band TX Range RX Range Tx/Rx Spacing GHz GHz GHz GHz Ceragon Proprietary and Confidential Page 119 of 225

120 Frequency Band TX Range RX Range Tx/Rx Spacing Ceragon Proprietary and Confidential Page 120 of 225

121 6.3.3 RFU-C Mechanical, Electrical, and Environmental Specifications RFU-C Mechanical, Electrical, and Environmental Specifications RFU-C RFU-Antenna Connection IDU-RFU Connection Polarization Standard Mounting OD Pole Operating Range Storage Transportation Power Consumption RFU-C 6-26 GHz Power Consumption RFU-C GHz Operating Temperature Height: 200 mm Width: 200 mm Depth: 85 mm Weight: 4kg/9 lbs Direct mount or remote using the same antenna type Remote mount: Standard flexible waveguide (frequency dependent) Coaxial cable RG-223 (100 m/300 ft), Belden 9914/RG-8 (300 m/1000 ft) or equivalent, N-type connectors (male) Vertical or Horizontal 50 mm-120 mm/2-4.5 (subject to vendor and antenna size) to -72 VDC ETS class T1.2, with a temperature range of -25 C to+85 C. ETS class 2.3, with a temperature range of -40 C to+85 C. 1+0: 22W 1+1: 39W 1+0: 26W 1+1: 43W Temperature range for continuous operating temperature with high reliability: -33 C to +55 C (-27 F to 131 F) Temperature range for exceptional temperatures; tested successfully, with limited margins: -45 C to +60 C (-49 F to 140 F) Relative Humidity 5% to 100% Ceragon Proprietary and Confidential Page 121 of 225

122 6.3.4 RFU-C Mediation Device Losses RFU-C Mediation Device Losses Configuration Interfaces 6-8 GHz 11 GHz GHz GHz GHz Flex WG Remote Mount antenna Added on remote mount configurations DirectMount Integrated antenna HSB Main Path Direct Mount with asymmetrical coupler Secondary Path DP (OMT) Direct Mount Integrated antenna HSB (OMT) Main Path Remote Mount with asymmetrical coupler Secondary Path /1+1 FD SP Integrated antenna DP (OMT) Remote Mount Notes: The antenna interface is always the RFU-C interface. If other antennas are to be used, an adaptor with a 0.1 db loss should be considered RFU-C Antenna Connection RFU-C uses Andrew, RFS, Xian Putian, Radio Wave, GD and Shenglu antennas. RFU-C can be mounted directly for all frequencies (6-42 GHz) using the following antenna types (for integrated antennas, specific antennas PNs are required): Andrew: VHLP series GD Radio Wave Xian Putian: WTG series Shenglu For remote mount installations, the following flexible waveguide flanges should be used (millimetric). The same antenna type (integrated) as indicated above can be used (recommended). Other antenna types using the flanges listed in the table below may be used. Ceragon Proprietary and Confidential Page 122 of 225

123 6.3.6 RFU-C Waveguide Flanges RFU-C Waveguide Flanges Frequency (GHz) Waveguide Standard Waveguide Flange Antenna Flange 6 WR137 PDR70 UDR70 7/8 WR112 PBR84 UBR84 10/11 WR90 PBR100 UBR WR75 PBR120 UBR WR62 PBR140 UBR WR42 PBR220 UBR WR28 PBR320 UBR WR22 UG383/U UG383/U If a different antenna type (CPR flange) is used, a flange adaptor is required. Please contact your Ceragon representative for details. For RFU-C transmit power specifications: RFU-C Transmit Power (dbm) For FRU-C receiver threshold specifications: RFU-C Receiver Threshold (RSL) BER = 10-6) 11 42GHz RFU-C is a roadmap item; parameters and availability are subject to change. Ceragon Proprietary and Confidential Page 123 of 225

124 HP/RFU-HP FibeAir 1500HP and RFU-HP are high transmit power RFUs designed for long haul applications with multiple carrier traffic. Together with their unique branching design, 1500HP/RFU-HP can chain up to five carriers per single antenna port and 10 carriers for dual port, making them ideal for Trunk or Multi Carrier applications. The 1500HP/RFU-HP can be installed in either indoor or outdoor configurations. The field-proven 1500HP/RFU-HP was designed to enable high quality wireless communication in the most cost-effective manner. With tens of thousands of units deployed worldwide, the 1500HP/RFU-HP serves mobile operators enabling them to reach over longer distances while enabling the use of smaller antennas. The RFU-HP also includes a power-saving feature ( green mode ) that enables the microwave system to automatically detect when link conditions allow it to use less power. 1500HP and RFU-HP 1RX support Space Diversity via Baseband Switching in the IDU (BBS). 1500HP 2RX, supports Space Diversity through IF Combining (IFC). Both types of Space Diversity are valid solutions to deal with the presence of multipath. Notes: Main Features of 1500HP/RFU-HP 12 Frequency range 1500 HP (11 GHz) 40 MHz bandwidth does not support IF Combining. For this frequency, Space Diversity is only available via BBS. 1500HP 2RX: 6-11GHz 1500HP 1RX: 6-11GHz RFU-HP: 6-8GHz Frequency accuracy ±4 ppm 13 Frequency source Synthesizer Installation type Split mount remote mount, all indoor (No direct mount) Diversity Optional innovative IF Combining Space Diversity for improved system gain (for 1500HP) 14, as well as BBS Space Diversity (all models) High transmit power Up to 33dBm in all indoor and split mount installations Configurable Modulation QPSK 256 QAM For guidance on the differences between 1500HP and RFU-HP, refer to RFU Selection Guide on page 110. Over temperature HP (11 GHz ) 40 MHz bandwidth does not support IF Combining. For this frequency, space diversity is only available via BBS. Ceragon Proprietary and Confidential Page 124 of 225

125 Configurable Channel Bandwidth 1500HP 2RX (6-11GHz): 10-30MHz 1500HP 1RX (6-11GHz): 10-30MHz 1500HP 1RX (11GHz wide): 24-40MHz RFU-HP 1RX (6-8GHz): MHz System Configurations Non-Protected (1+0), Protected (1+1), Space Diversity, 2+0/2+2 XPIC, N+0, N+1 Variety of interfaces for TDM and IP XPIC and CCDP Built-in XPIC (Cross Polarization Interference Canceller) and Co-Channel Dual Polarization (CCDP) feature for double transmission capacity, and more bandwidth efficiency Power Saving Mode option - Enables the microwave system to automatically detect when link conditions allow it to use less power (for RFU-HP) Tx Range (Manual/ATPC) Up to 20dB dynamic range ATPC (Automatic Tx Power Control) RF Channel Selection Via EMS/NMS NEBS Level 3 NEBS compliance Ceragon Proprietary and Confidential Page 125 of 225

126 HP/RFU-HP Frequency Bands The frequency band of each radio is listed in the following table. Frequency Band Frequency Range (GHz) Channel Bandwidth L6 GHz to /56MHz U6 GHz to MHz to 40/56 /60 MHz to MHz to 28/56 MHz 7 GHz to /56 MHz to /56 MHz to MHz 8 GHz to MHz to 28/56 MHz to MHz to 28/56 MHz 11 GHz to MHz to 40/56 Ceragon Proprietary and Confidential Page 126 of 225

127 HP/RFU-HP Mechanical, Electrical, and Environmental Specifications 1500HP/RFU-HP Mechanical, Electrical, and Environmental Specifications Transceiver (RFU) Dimensions OCB Branching (Split Mount and Compact All-Indoor ) IDU-RFU Connection Height: 490 mm (19 ) Width: 144 mm (6 ) Depth: 280 mm (11 ) Weight: 7 kg (15 lbs) (excluding Branching) Height: 420 mm (19 ) Width: 110 mm (6 ) Depth: 380 mm (11 ) Weight: 7 kg (15 lbs) (excluding Branching) Recommended torque for RFU-OCB connection: 17 Nm Coaxial cable RG-223 (100 m/300 ft), Belden 9914/RG-8 (300 m/1000 ft) or equivalent, N-type connectors (male) RFU Power Consumption Split Mount (29dBm): 80W All indoor (32dBm) : 100W Storage Transportation Power Supply Operating Temperature ETS class T1.2, with a temperature range of -25 C to+85 C. ETS class 2.3, with a temperature range of -40 C to+85 C to -72 VDC Temperature range for continuous operating temperature with high reliability: -33 C to +55 C (-27 F to 131 F) Temperature range for exceptional temperatures; tested successfully, with limited margins: -45 C to +60 C (-49 F to 140 F) Relative Humidity 5% to 100% For additional information: Power Consumption with RFU-HP in Power Saving Mode Ceragon Proprietary and Confidential Page 127 of 225

128 HP/RFU-HP Functional Block Diagram and Concept of Operation The RFU handles RF signal processing. The RFU encompasses the RF transmitter and receiver with all their related functions. The 1500HP/RFU-HP product line was designed to answer the need for a high power RF module together with IF combining functionality and the ability to concatenate several carriers with minimal RF branching loss. This section briefly describes the basic block diagrams for the various types of RFUs included in the 1500HP/RFU-HP product line. Figure 1: 1500HP 2RX in 1+0 SD Configuration VCO TX Board OCB Antenna main 350MHz IF TX TX chain Pre- Amp PA TX Quadplexer FSK -48V 140MHz Controller and peripherals PSU C o n n e c t o r C o n n e c t o r 10M DC / CTRL combiner TCXO RX chain RX chain LNA LNA RF LPBK RX Main RX Diversity RX RX Extention port diplexer IF & controller Board XLO XPIC SW VCO RX Antenna Diversity Chassis IDU XPIC source (Ntype conn.) sharing \ RSL ind. (TNC conn.) Figure 2: 1500HP 1RX in 1+0 SD Configuration VCO TX Board OCB Antenna main 350MHz IF TX TX chain FMM FLM TX Quadplexer FSK -48V 140MHz Controller and peripherals PSU C o n n e c t o r C o n n e c t o r DC / CTRL TXCO RX chain LNA RF LPBK RX Main RX Extention port 10M diplexer XLO XPIC SW VCO IF & controller Board RX Board Chassis IDU XPIC source (Ntype conn.) sharing \ RSL ind. (TNC conn.) Ceragon Proprietary and Confidential Page 128 of 225

129 Figure 3: RFU-HP 1RX in 1+0 SD Configuration VCO OCB Antenna main IDU (BMA conn.) XPIC source sharing \ RSL ind. (BMA conn.) 350MHz FSK Quadplexer -48V 140MHz Controller and peripherals PSU section C o n n e c t o r C o n n e c t o r IF RX RFIC TX RFIC DC / CTRL 40M TX chain RX chain Pre- Amp LNA PA RF LPBK TX RX Extention port 40M diplexer XLO XPIC SW VCO PSC Chassis TRX XPIC source sharing \ RSL ind. (TNC conn.) Each of these RFU types must be connected to an OCB (Outdoor Circulator Block) which serves as both a narrow diplexer and a mediation device to facilitate antenna connection. For additional information: 1500HP/RFU-HP OCBs Ceragon Proprietary and Confidential Page 129 of 225

130 HP/RFU-HP Comparison Table The following table summarizes the differences between the 1500HP 2RX and 1RX and the RFU-HP. 1500HP/RFU-HP Comparison Table Feature 1500HP 2RX 1500HP 1RX RFU-HP 1RX Notes Frequency Bands Support 6L,6H,7,8,11GHz 6L,6H,7,8,11GHz 6L,6H,7,8GHz 3.5MHz 56 MHz MHz 30 MHz 40MHz -- ** ** 11GHz only supports 24-40MHz channels only Split-Mount All are compatible with OCBs from both generations All-Indoor All are compatible with ICBs Space Diversity BBS and IFC 15 BBS BBS IFC - IF Combining BBS - Base Band Switching Frequency Diversity 1+0/2+0/1+1/2+2 N+1 N+0 ( N>2) High Power Only the RFU-HP has the same power for split mount and all indoor installation. Refer to 1500HP/RFU- HP Models and Part Numbers on page 154. Direct Mount Antenna Power Saving Mode Power consumption changes with TX power Note that the main differences between the 1500HP 1RX and RFU-HP 1RX are: RFU-HP offers higher TX power for split mount The RFU-HP 1RX offers full support for 3.5M-56MHz channels. The RFU-HP 1RX supports the green-mode feature Both systems are fully compatible with all OCB and ICB devices HP (11 GHz ) 40 MHz bandwidth does not support IF Combining. For this frequency, space diversity is only available via BBS. Ceragon Proprietary and Confidential Page 130 of 225

131 HP/RFU-HP System Configurations Split Mount and All indoor The 1500HP/RFU-HP radios can be installed either in split mount or in all indoor configurations. The following configurations are applicable for Split-Mount or all indoor installations: Notes: Unprotected N to 10+0 Data is transmitted through N channels, without redundancy (protection) Hot Standby HSB, 2+2 HSB Two RFUs use the same RF channel connected via a coupler. One channel transmits (Active) and the other acts as a backup (Standby). A 2+2 HSB configuration uses two RFUs which are chained using two frequencies and connected via a coupler to the other pair of RFUs. N+1 Frequency Diversity - N+1 (1+1 to 9+1) Data is transmitted through N channels and an additional (+1) frequency channel, which protects the N channels. If failure or signal degradation occurs in one of the N channels, the +1 channel carries the data of the affected N carrier. Additional configurations, such as 14+2, can be achieved using two racks. Space Diversity can be used in each of the configurations. When using BBS for SD (1500HP 1RX/RFU-HP), ACM is not supported. When the 1500HP/RFU-HP is mounted in a Split-Mount configuration, up to five RFUs can be chained on one pole mount (the total is ten RFUs for a dual pole antenna). When the 1500HP/RFU-HP is installed in an All Indoor configuration, there are several installation options: In ETSI rack up to ten radio carriers per rack In 19 open rack up to five radio carriers per subrack Compact assembly up to two radio carriers in horizontal placement (without a subrack) Two types of branching options are available for all indoor configurations: Using ICBs Vertical assembly, up to 10 carriers per rack (five carriers per subrack) Using OCBs Compact horizontal assembly, up to 2 carriers per subrack HP/RFU-HP Space Diversity Support In long distance wireless links, multipath phenomenon commonly exist, whereby fading occurs over time, space, and frequency. The 1500HP RFU provides two types of Space Diversity optimizations, which are ideal solutions for the multipath phenomenon: IF Combining (IFC) BBS (Base Band Switching) The RFU-HP supports BBS Space Diversity, but not IFC. Ceragon Proprietary and Confidential Page 131 of 225

132 Space Diversity with Multiple RFUs Space Diversity with Single RFU IF Combining (IFC) Mechanism FibeAir 1500HP includes an IF combining mechanism, which uses an innovative digital optimization algorithm to combine the signals received from both antennas in order to improve signal quality. When distortion occurs, it is measured in both receiver paths, and a new combined signal is produced. This can improve the system gain by up to 3 db. IFC Space Diversity can be used with single and multiple RFUs. A delay calibration for the diversity waveguide is required and is performed automatically via the NMS. Each 1500HP has built-in IFC Space Diversity functionality, with one transmitter and two receivers. The receivers receive two different signals from two antennas, which are installed meters apart. There are two options for connecting the RFUs to the diversity antennas: Waveguide to coaxial cable Uses a waveguide adaptor (CPR type) connected to an N-type coaxial cable. This is the default option. Elliptical waveguide Uses a waveguide connector (CPR type) with an elliptical waveguide Baseband Switching (BBS) Both FibeAir 1500HP and FibeAir RFU-HP support BBS Space Diversity. In this option, there are two RFUs instead of a single RFU with two receivers. The actual BBS Space Diversity switching is performed in the IDU. The modem switches to the other RF signal when interference occurs, and returns to the main signal when the interference is gone. In this way, the system performs optimum signal receiving by using the signal that provides the best performance. Note: When using BBS for SD (1500HP 1RX/RFU-HP), ACM is not supported Ceragon Proprietary and Confidential Page 132 of 225

133 6.4.8 Split Mount Configuration and Branching Network For multiple carriers, up to five carriers can be cascaded and circulated together to the antenna port. Branching networks are the units which perform this function and route the signals from the RFUs to the antenna. The branching network can contain multiple OCBs or ICBs. When using a Split-Mount or All-Indoor compact (horizontal) configuration, the OCB branching network is used. When using an All-Indoor vertical configuration, the ICB branching network is used. The main differences in branching concept between the OCB and the ICB relate to how the signals are circulated. OCB The Tx and the Rx path circulate together to the main OCB port. When chaining multiple OCBs, each Tx signal is chained to the OCB Rx signal and so on (uses S-bend section). For more details, refer to 1500HP/RFU-HP OCBs on page 134. ICB All the Tx signals are chained together to one Tx port (at the ICC) and all the Rx signals are chained together to one Rx port (at the ICC). The ICC circulates all the Tx and the Rx signals to one antenna port. For more information, refer to Indoor Circulator Block (ICB) on page 141. All-Indoor Vertical Branching Split-Mount Branching and All-Indoor Compact Ceragon Proprietary and Confidential Page 133 of 225

134 HP/RFU-HP OCBs The OCB (Outdoor Circulator Block) has the following main purposes: Hosts the circulators and the attached filters. Routes the RF signal in the correct direction, through the filters and circulators. Enables RFU connection to the Main and Diversity antennas. FibeAir 1500HP and RFU-HP supports two types of OCBs: OCB (Older Type) New OCB Old OCB New OCB Old OCB The Older Type OCB has two types, Type 1 and Type 2. The difference between the two types is the circulator direction. Depending on the configuration, OCB Type 1 or Type2 is used together with waveguide shorts, loads, U Bends, or couplers. Each OCB has four waveguide access points: two in the front, and two at the rear. The diversity access point is optional. If the system is not configured for diversity, all the relevant access points on the OCB must be terminated using waveguide shorts. The two OCB types (with and without IFC Space Diversity) have different part numbers. Ceragon Proprietary and Confidential Page 134 of 225

135 The following block diagrams show the difference between the two OCBs and the additional Diversity Circ block which is added in some diversity configurations. Old OCB Type 1 Old OCB Type 1 and Type 2 Description Ceragon Proprietary and Confidential Page 135 of 225

136 New OCB The new OCB is optimized for configurations that do not use IFC Space Diversity. To support IFC Space Diversity, a diversity block is added. The new OCB has only one type, and can be connected to an antenna via a flexible waveguide. The new OCB connection is at the rear of the OCB. It includes proprietary accessories (different than those used for the older OCB). Each OCB has three waveguide access points: The In/Out port is located at the rear of the OCB. The OCB ports include: Tx port Rx Port Diversity port If the system is not configured for diversity, all the relevant access points on the OCB must be terminated using waveguide shorts. Unused Rx ports are terminated with a 50 ohm termination. New OCB and DCB Block Diagram New OCB components include the following: RF Filters RF Filters are used for specific frequency channels and Tx/Rx separation. The filters are attached to the OCB, and each RFU contains one Rx and one Tx filter. In an IFC Space Diversity configuration, each RFU contains two Rx filters (which combine the IF signals) and one Tx filter. The filters can be replaced without removing the OCB. Ceragon Proprietary and Confidential Page 136 of 225

137 DCB (Diversity Circulator Block) THE DCB is an external block which is added in IFC Space Diversity configurations. The DCB is connected to the diversity port and can chain two OCBs. Coupler Kit The coupler kit is used for 1+1 Hot Standby (HSB) configurations. U Bend The U Bend connects the chained DCB (Diversity Circulator Block) in N+1/N+0 configurations. S Bend The S Bend connects the chained OCB (Outdoor Circulator Block) in N+1 /N+ 0 configurations. Pole Mount Kit The Pole Mount Kit can fasten up to five OCBs and the RFUs to the pole. The kit enables fast and easy pole mount installation New OCB Component Summary New OCB Component Summary Component Name Marketing Model Marketing Description Picture DCB DCBf DCB Diversity Block f GHz kit CPLR OCB-CPLR-f OCB Coupler f GHz CPLR Sym OCB-CPLR_SYM-f OCB symmetrical Coupler fghz U Bend DCB-UBend DCB Ubend connection f GHz S Bend OCB-SBend OCB SBend connection f GHz Pole Mount OCB-Pole Mount OCB-Pole Mount Note: f= 6L, 6H, 7, 8, 11 GHz Ceragon Proprietary and Confidential Page 137 of 225

138 6.4.9 Split-Mount Branching Loss When designing a link budget calculation, the branching loss (db) should be considered as per specific configuration. This section contains tables that list the branching loss for the following Split-Mount configurations. Interfaces FD/ CCDP with DP Antenna SP Non-adjacent Channels 0 (1c) 0 (1c) 0.5 (2c) 0.5 (2c) 1.0 (3c) 1.0 (3c) 1.5 (4c) 1.5 (4c) 2 (5c) 2 (6c) 0 (1c) 0.5 (2c) 1.0 (3c) 1.5 (4c) 2.0 (5c) NA NA NA NA NA Notes: (c) Radio Carrier CCDP Co-channel dual polarization SP Single pole antenna DP Dual pole antenna In addition, the following losses will be added when using these items: Item Where to Use Loss (db) Flex WG All configurations m Coax cable Diversity path 6-8/11 GHz 5/6.5 Symmetrical Coupler Adjacent channel configuration. 3.5 Asymmetrical coupler 1+1 HSB configurations Main: 1.6 Coupled: Upgrade Procedure The following components need to be added when upgrading from a 1+0 to an N+1 Split-Mount configuration: OCBs RFUs IDU/IDMs Flexible waveguides When adding RF channels or carriers, RFUs and OCBs with specific filters need to be added as well. The OCBs are chained together using couplers (for the same frequency) or U bends/s bends (for different frequencies), in accordance with the specific configuration. Open ports on the OCBs are terminated with 50 ohm terminations. Detailed upgrade procedure documents are available for specific configurations. Please note that legacy OCBs can be upgraded and cascaded with the new OCB. Please contact your Ceragon representative for details. Ceragon Proprietary and Confidential Page 138 of 225

139 HP/RFU-HP All Indoor Configurations and Branching Network All-Indoor configurations are when all the equipment is installed indoors (room, shelter) and an elliptical waveguide connects the radio output port from the room to the antenna. A basic block diagram for a trunk system, including the main blocks, is shown in the following figure. The block diagram includes marked interface points which shall serve as reference points for several technical parameters used in this document. Block Diagram of Trunk System All-Indoor System with Five IP-10 Carriers Ceragon Proprietary and Confidential Page 139 of 225

140 All-Indoor System with Ten IP-10 Carriers The branching concept (as described in Split Mount Configuration and Branching Network on page 133) is similar to All-Indoor application. When using All-Indoor configurations, there are two types of branching implementations: Using ICBs Vertical assembly, up to 10 carriers per rack (five carriers per subrack). Using NEW OCBs Compact horizontal assembly, up to two carriers per subrack. All-Indoor Installations Ceragon Proprietary and Confidential Page 140 of 225

141 RFU Subrack Components Subrack for ETSI Rack Subrack The subrack hosts all the RFU components and connections, as shown in the previous figure. The subrack includes up to five RFUs per subrack (each RFU connects to an ICB). RFU with Branching Indoor Circulator Block (ICB) Each RFU is connected to one ICB, and several ICBs are chained to each other. The chained ICBs carry different RF channels and are connected to a single ICC, which sums the RF signals. The main ICB functions include: Ceragon Proprietary and Confidential Page 141 of 225

142 Hosts the circulators and filters. Routes the RF signals in the correct direction, via the filters and circulators. The ICB is a modular standalone unit. When system expansion is necessary, additional ICBs are added and chained with the existing ICBs. The branching chain to neighbor ICB goes through the holes at the side. A long screw connects the ICBs to each other and the last ICB at the chain is terminated with a 50ohm termination, as shown below. Note: The diversity port does not need to be terminated if the diversity filter is not attached to the ICB. ICB Branching Chain RF Filters The RF Filters are used for specific frequency channels and Tx/Rx separation. The filters are attached to the ICB, and each RFU contains one Rx and one Tx filter. In an IFC Space Diversity configuration, each RFU contains two Rx filters to combine the IF signals, along with one Tx filter. Ceragon Proprietary and Confidential Page 142 of 225

143 Indoor Combiner Circulator (ICC) The ICC does not perform space diversity ICB summing (single output port). ICC The ICC sums the Rx and Tx signals and combines the N channels to the output ports (one or two, in accordance with the configuration). Indoor Combiner Circulator Diversity (ICCD) The ICCD performs space diversity ICB summing (two output ports). ICCD Patch Panel The ICB s IF and XPIC cables are connected to the patch panel. The IDU s IF cables are connected to the specific RFU location. An XPIC cable is used between two RFUs which are using the same Tx and Rx filters with different polarizations (V and H). Ceragon Proprietary and Confidential Page 143 of 225

144 Fan Tray The fan tray contains eight controlled and monitored fans, which cool the RFU heat dissipations. The fan tray is a tray which is part of ETSI rack (as shown above), while when using a 19 frame rack a fan tray is a separate unit which must be assembled separately (shown below). Fan Tray in 19 Frame Rack Rigid Waveguides - T12, T13 and T14 Rigid waveguide sections are assembled in the rack to connect the ICC/ICCD from the bottom to the top of the rack (C ). The specific Rigid WG sections to be used depend on the configuration. T12 Rigid Waveguide T13 Rigid Waveguide Ceragon Proprietary and Confidential Page 144 of 225

145 All-Indoor Configuration Example In this configuration, three ICBs are chained together and connected to a vertical ICC, and two ICBs are chained together and connected to a horizontal ICC polarization. The RF components include: Five RFUs Five ICBs Two ICCs 4+1 XPIC Assembly Configuration Additional Assembly Configuration Examples Ceragon Proprietary and Confidential Page 145 of 225

146 All-Indoor Rack Types Three types of racks can be used in an all-indoor configuration: 19 lab rack ( open frame ) 19 rack ETSI rack The 19 rack is not commonly used in Ceragon configurations. The 19 lab rack (open frame) contains a subrack that is preassembled at the factory and then shipped. The customer can also use an existing rack and the subrack is installed separately at the site Rack Type Examples Lab Rack (Open Frame) Examples Ceragon Proprietary and Confidential Page 146 of 225

147 19 Rack Example ETSI Rack Example Ceragon Proprietary and Confidential Page 147 of 225

148 When a configuration includes more than ten carriers, two racks are assembled and connected. Configuration with More than Ten Carriers Two Connected Racks Ceragon Proprietary and Confidential Page 148 of 225

149 All-Indoor Branching Loss ICC has a 0 db loss, since the RFU is calibrated to Pmax, together with the filter and 1+0 branching loss. The following table presents the branching loss per configuration and the Elliptical wave guide (WG) losses per meter which will be add for each installation (dependant on the WG length). Configuration Interfaces FD L 4 WG losses per 100m 6H 4.5 7/8GHz 6 All-Indoor Configuration All-Indoor Symmetrical Coupler CCDP with DP antenna SP Non adjacent channels CCDP with DP antenna Upgrade Ready Interfaces WG losses per 100m Symmetrical Coupler CCDP with DP antenna 11GHz 10 Added to adjacent channel configuration 3 Tx and Rx 0.3 (1c) 0.3 (1c) 0.7 (2c) 0.7 (2c) 1.1 (3c) Diversity RX 0.2 (1c) 0.2 (1c) 0.6 (2c) 0.6 (2c) 1.0 (3c) Tx and Rx 0.3 (1c) 0.7 (2c) 1.1 (3c) 1.5 (4c) 1.9 (5c) Diversity RX 0.2 (1c) 0.6 (2c) 1.0 (3c) 1.4 (4c) 1.8 (5c) Tx and Rx 0.3 (1c) 0.7 (1c) 1.1 (2c) 1.1 (2c) 1.5 (3c) Diversity RX 0.2 (1c) 0.6 (1c) 1.0 (2c) 1.0 (2c) 1.4 (3c) 6L H 4.5 7/8GHz 6 11GHz 10 Added to adjacent channel configuration Tx and Rx 1.5 (3c) 1.9 (4c) 1.9 (4c) 2.3 (5c) 2.3 (6c) Diversity RX 1.4 (3c) 1.8 (4c) 1.8 (4c) 2.2 (5c) 2.2 (6c) SP Non adjacent channels Tx and Rx Diversity RX NA NA NA NA NA CCDP with DP antenna Upgrade Ready Tx and Rx 1.5 (3c) 1.9 (4c) 1.9 (4c) 2.3 (5c) 2.3 (6c) Diversity RX 1.4 (3c) 1.8 (4c) 1.8 (4c) 2.2 (5c) 2.2 (6c) Ceragon Proprietary and Confidential Page 149 of 225

150 HP/RFU-HP All Indoor Compact (Horizontal) For minimal rack space usage, an All-Indoor configuration can be installed in horizontal position using the new OCB in a 19 rack or ETSI open rack/ frame rack. The New OCB is compliant with NEBS GR-1089-CORE, GR-63-CORE standards. Note: This installation type and configuration does not require a fan tray. This installation type is compatible with the following RFUs PN: Non Space Diversity All-Indoor 15HPA-1R-RFU-f 15HPA-2R-RFU-f 15HPA-1R-RFU-11w 1500HP RFU All-Indoor 1Rx RF Unit 1500HP RFU All-Indoor Space Diversity Ceragon Proprietary and Confidential Page 150 of 225

151 1500HP RFU All-Indoor 1Rx RF Unit, 11G 40MHz Main Configurations East West East West 1+1 HSB Compact Front View 1+1 HSB Compact Rear View Ceragon Proprietary and Confidential Page 151 of 225

152 All-Indoor Compact (Horizontal) Placements Components The following table lists the components for All-Indoor compact placements: All-Indoor Compact Placement Components Component Name Marketing Model Marketing Description Picture DCB DCBf DCB Diversity Block f GHz kit CPLR OCB-CPLR-f OCB Coupler f GHz SBend OCB-SBend OCB SBend Connection f GHz Rack Adapter OCB 19 Rack Adapt OCB-Pole Mount Rack Adapter OCB ETSI Rack Adapt OCB-Pole Mount Note: f= 6L, 6H, 7, 8, 11 GHz Power Distribution Unit (PDU) The PDU distributes the power supply (-48V) from the main power input to the relevant IDU. The PDU is preassembled and wired in an ETSI rack and is provided separately, when required, for a 19 lab rack. When ordering a 19 configuration, there are two rack assembly options: 19 lab rack provided separately 19 lab rack provided by the customer For both options, a PDU for 19 can be provided upon request. There are two types of PDU. The default PDU which has been assembled with each ETSI rack contains: Two main switches one for each five IDU carriers Two FAN tray switches Ceragon Proprietary and Confidential Page 152 of 225

153 1A. The default PDU which is assembled with the ETSI rack has a special addition of a plastic cover. For special cases, when PDU protection is required, a PDU with plastic protection cover can be provided. The PN for this PDU with protection cover is: 32T-PDU_CVR. A PDU which distributes 10 x DC signals, the PDU type can be preassembled with an ETSI Rack and needs to be specially ordered because it is not the default PDU. PDU with 10 Switches PN: 32T-PDU10 Ceragon Proprietary and Confidential Page 153 of 225

154 HP/RFU-HP Models and Part Numbers The following table lists and describes the available 1500HP/RFU-HP models. RFU Models Marketing Model 15HP-RFU-7 15HP-RFU-8 15HP-RFU-6L 15HP-RFU-6H 15HP-RFU-11 Description 1500HP 7G 2RX SM / All Indoor 1500HP 8G 2RX SM / All Indoor 1500HP 6LG 2RX SM / All Indoor 1500HP 6HG 2RX SM / All Indoor 1500HP 11G 2RX SM / All Indoor 15HPS-1R-RFU-7 15HPS-1R-RFU-8 15HPS-1R-RFU-6L 15HPS-1R-RFU-6H 15HPS-1R-RFU-11 15HPS-1R-RFU-11w 1500HP 7G 1RX SM 1500HP 8G 1RX SM 1500HP 6LG 1RX SM 1500HP 6HG 1RX SM 1500HP 11G 1RX SM 1500HP 11G 1RX SM 40M (24-40MHz channels) 15HPA-1R-RFU-7 15HPA-1R-RFU-8 15HPA-1R-RFU-6L 15HPA-1R-RFU-6H 15HPA-1R-RFU HP 7G 1RX All Indoor 1500HP 8G 1RX All Indoor 1500HP 6LG 1RX All Indoor 1500HP 6HG 1RX All Indoor 1500HP 11G 1RX All Indoor 15HPA-2R-RFU-7 15HPA-2R-RFU-8 15HPA-2R-RFU-6L 15HPA-2R-RFU-6H 15HPA-2R-RFU HP 7G 2RX All Indoor 1500HP 8G 2RX All Indoor 1500HP 6LG 2RX All Indoor 1500HP 6HG 2RX All Indoor 1500HP 11G 2RX All Indoor RFU-HP-1R-6H RFU-HP-1R-6L RFU-HP-1R-7 RFU-HP-1R-8 RFU-HP 6HG 1Rx up to 56M SM / All Indoor RFU-HP 6LG 1Rx up to 56M SM / All Indoor RFU-HP 7G 1Rx up to 56M SM / All Indoor RFU-HP 8G 1Rx up to 56M SM / All Indoor Ceragon Proprietary and Confidential Page 154 of 225

155 OCB Part Numbers The following table presents the various RFU options and the configurations in which they are used. OCB Part Numbers Diversity/Non-Diversity Space Diversity IFC (2Rx) (6, 7, 8,11GHz) Non Space Diversity (1Rx) (6, 7, 8GHz) 11GHz Non Space Diversity (1Rx) 16 Split-Mount 15OCBf-SD-xxxy-ZZZ-H/L 15OCBf-xxxy-ZZ-H/L 15OCB11w-xxxy-ZZ-H/L OCB Part Numbers for All Indoor Compact Diversity/Non-Diversity Space Diversity IFC (2Rx) (6, 7,8 GHz) Space Diversity IFC (2Rx) (11GHz) Non Space Diversity (1Rx) (6, 7,8GHz) 11GHz Non Space Diversity (1Rx) 17 All Indoor Compact 15OCBf-SD-xxxy-ZZ-H/L 15OCB11w-SD-xxxy-ZZ-H/L 15OCBf-xxxy-ZZ-H/L 15OCB11w-xxxy-ZZ-H/L OCB Part Number Format Place Holder in Marketing Model f Possible Values 6L,6H,7,8,11 Description and Remarks xxx [MHz] TRS in MHz Y A Z Ceragon TRS block designation ZZZ Examples: 1W3 Wide filters covering channels Only channel 03, 28MHz channel MHz Narrow filters allowing concatenation using OCBs covering channels 3 and 4. Designation of the channels the OCB is covering H/L H or L Designating TX High and TX low GHz OCB is a wide BW OCB which supports up to 40MHz, while the other OCBs (6L, 6H, 7, 8GHz) support up to 30MHz. 11GHz OCB is a wide BW OCB which supports up to 40MHz, while the other OCBs (6L, 6H, 7, 8GHz) support up to 30MHz. Ceragon Proprietary and Confidential Page 155 of 225

156 Generic All-Indoor Configurations Part Numbers The following tables contain a list of typical All-Indoor configurations. All-Indoor Configurations (1+0 /1+1 HSB) 1+0 / 1+1 HSB 32T-f_1+0 32T-f_1+0_EW 32T-f_1+0_SD 32T-f_1+0_SD_EW 32T-f_1+1_HSB 32T-f_1+1_HSB_EW 32T-f_1+1_HSB_SD 32T-f_1+1_HSB_SD_EW 3200T-f_ T-f_1+0_East West 3200T-f_1+0_Space Diversity 3200T-f_1+0_Space Diversity East West 3200T-f_1+1_HSB 3200T-f_1+1_HSB_East West 3200T-f_1+1_HSB_Space Diversity 3200T-f_1+1_HSB_Space Diversity East West All-Indoor Configurations (N+0/N+1 XPIC) N+0 / N+1 XPIC 32T-f_1+1/2+0_X 32T-f_2+1/3+0_X 32T-f_3+1/4+0_X 32T-f_4+1/5+0_X 32T-f_5+1/6+0_X 32T-f_6+1/7+0_X 32T-f_7+1/8+0_X 32T-f_8+1/9+0_X 32T-f_9+1/10+0_X 3200T-f_1+1/2+0 XPIC 3200T-f_2+1/3+0 XPIC 3200T-f_3+1/4+0 XPIC 3200T-f_4+1/5+0 XPIC 3200T-f_5+1/6+0 XPIC 3200T-f_6+1/7+0 XPIC 3200T-f_7+1/8+0 XPIC 3200T-f_8+1/9+0 XPIC 3200T-f_9+1/10+0 XPIC Ceragon Proprietary and Confidential Page 156 of 225

157 All-Indoor Configurations (N+0 / N+1 XPIC Space Diversity) N+0 / N+1 XPIC Space Diversity 32T-f_1+1/2+0_X _SD 32T-f_2+1/3+0_X _SD 32T-f_3+1/4+0_X_SD 32T-f_4+1/5+0_X_SD 32T-f_5+1/6+0_X_SD 32T-f_6+1/7+0_X_SD 32T-f_7+1/8+0_X_SD 32T-f_8+1/9+0_X_SD 3200T-f_1+1/2+0 XPIC Space Diversity 3200T-f_2+1/3+0 XPIC Space Diversity 3200T-f_3+1/4+0 XPIC Space Diversity 3200T-f_4+1/5+0 XPIC Space Diversity 3200T-f_5+1/6+0 XPIC Space Diversity 3200T-f_6+1/7+0 XPIC Space Diversity 3200T-f_7+1/8+0 XPIC Space Diversity 3200T-f_8+1/9+0 XPIC Space Diversity All-Indoor Configurations (N+0 / N+1 XPIC Space Diversity) N+0 / N+1 XPIC Space Diversity 32T-f_1+1/2+0_X _SD 32T-f_2+1/3+0_X _SD 32T-f_3+1/4+0_X_SD 32T-f_4+1/5+0_X_SD 32T-f_5+1/6+0_X_SD 32T-f_6+1/7+0_X_SD 32T-f_7+1/8+0_X_SD 32T-f_8+1/9+0_X_SD 32T-f_9+1/10+0_X_SD 32T-f_1+1/2+0_X_EW 32T-f_2+1/3+0_X_EW 32T-f_3+1/4+0_X_EW 32T-f_4+1/5+0_X_EW 32T-f_1+1/2+0_X_SD_EW 32T-f_2+1/3+0_X_SD_EW 32T-f_3+1/4+0_X_SD_EW 32T-f_4+1/5+0_X_SD_EW 3200T-f_1+1/2+0 XPIC Space Diversity 3200T-f_2+1/3+0 XPIC Space Diversity 3200T-f_3+1/4+0 XPIC Space Diversity 3200T-f_4+1/5+0 XPIC Space Diversity 3200T-f_5+1/6+0 XPIC Space Diversity 3200T-f_6+1/7+0 XPIC Space Diversity 3200T-f_7+1/8+0 XPIC Space Diversity 3200T-f_8+1/9+0 XPIC Space Diversity 3200T-f_9+1/10+0 XPIC Space Diversity 3200T-f_1+1/2+0 XPIC East West 3200T-f_2+1/3+0 XPIC East West 3200T-f_3+1/4+0 XPIC East West 3200T-f_4+1/5+0 XPIC East West 3200T-f_1+1/2+0 XPIC East West Space Diversity 3200T-f_2+1/3+0 XPIC East West Space Diversity 3200T-f_3+1/4+0 XPIC East West Space Diversity 3200T-f_4+1/5+0 XPIC East West Space Diversity Ceragon Proprietary and Confidential Page 157 of 225

158 N+0/N+1 Single Pol All-Indoor Configurations (N+0/N+1 Single Pol) 32T-f_1+1/2+0_SP 32T-f_2+1/3+0_SP 32T-f_3+1/4+0_SP 32T-f_4+1/5+0_SP 3200T-f_1+1/2+0_SP 3200T-f_2+1/3+0_SP 3200T-f_3+1/4+0_SP 3200T-f_4+1/5+0_SP All-Indoor Configurations (N+0/N+1 Single Pol Space Diversity) N+0/N+1 Single Pol Space Diversity 32T-f_1+1/2+0_SP_SD 32T-f_2+1/3+0_SP_SD 32T-f_3+1/4+0_SP_SD 32T-f_4+1/5+0_SP_SD 32T-f_1+1/2+0_SP_EW 32T-f_2+1/3+0_SP_EW 32T-f_3+1/4+0_SP_EW 32T-f_4+1/5+0_SP_EW 32T-f_1+1/2+0_SP_SD_EW 32T-f_2+1/3+0_SP_SD_EW 32T-f_3+1/4+0_SP_SD_EW 32T-f_4+1/5+0_SP_EW 3200T-f_1+1/2+0_Single Pole Space Diversity 3200T-f_2+1/3+0_Single Pole Space Diversity 3200T-f_3+1/4+0_Single Pole Space Diversity 3200T-f_4+1/5+0_Single Pole Space Diversity 3200T-f_1+1/2+0_Single Pole East West 3200T-f_2+1/3+0_Single Pole East West 3200T-f_3+1/4+0_Single Pole East West 3200T-f_4+1/5+0_Single Pole East West 3200T-f_1+1/2+0_Single Pole Space Diversity East West 3200T-f_2+1/3+0_Single Pole Space Diversity East West 3200T-f_3+1/4+0_Single Pole Space Diversity East West 3200T-f_4+1/5+0_Single Pole East West All-Indoor Configurations (N+0/N+1 XPIC Upgrade ready) N+0/N+1 XPIC Upgrade Ready 32T-f_1+1/2+0_X_UR 32T-f_2+1/3+0_X_UR 32T-f_3+1/4+0_X_UR 32T-f_4+1/5+0_X_UR 3200T-f_1+1/2+0_XPIC_Upgrade Ready 3200T-f_2+1/3+0_XPIC_Upgrade Ready 3200T-f_3+1/4+0_XPIC_Upgrade Ready 3200T-f_4+1/5+0_XPIC_Upgrade Ready Ceragon Proprietary and Confidential Page 158 of 225

159 All-Indoor Configurations (N+0/N+1 XPIC Space Diversity Upgrade-Ready) N+0/N+1 XPIC Space Diversity Upgrade Ready 32T-f_1+1/2+0_X_SD_UR 32T-f_2+1/3+0_X_SD_UR 32T-f_3+1/4+0_X_SD_UR 32T-f_4+1/5+0_X_SD_UR 3200T-f_1+1/2+0_XPIC_Space Diversity Upgrade Ready 3200T-f_2+1/3+0_XPIC_Space Diversity Upgrade Ready 3200T-f_3+1/4+0_XPIC_Space Diversity Upgrade Ready 3200T-f_4+1/5+0_XPIC_Space Diversity Upgrade Ready 19" Without Rack All-Indoor Configurations (19" Without Rack) 32T19-f_1+0_WO_rack 32T19-f_1+0_EW_WO_rack 32T19-f_1+0_SD_WO_rack 32T19-f_1+0_SD_EW_WO_rack 32T19-f_1+1_HSB_WO_rack 32T19-f_1+1_HSB_SD_WO_rack 32T19-f_1+1_HSB_EW_WO_rack 3200T19_inch-f_1+0_Without_rack 3200T19_inch-f_1+0_East West Without rack 3200T19_inch-f_1+0_Space Diversity Without rack 3200T19_inch-f_1+0_Space Diversity East West Without rack 3200T19_inch-f_1+1_HSB_Without_rack 3200T19_inch-f_1+1_HSB_Space Diversity Without rack 3200T19_inch-f_1+1_HSB_East West Without rack 32T19-f_1+1_HSB_SD_EW_WO_rack 3200T19_inch-f_1+1_HSB_Space Diversity East West Without rack For additional configurations and details, please contact your Ceragon representative. For 1500HP/RFU-HP transmit power specifications 1500HP/RFU-HP Transmit Power (dbm) For 1500HP/RFU-HP receiver threshold specifications: 1500HP/RFU-HP Receiver Threshold (RSL) = 10-6) Ceragon Proprietary and Confidential Page 159 of 225

160 6.5 RFH-HS FibeAir RFU-HS is a high transmit power RFU for long-haul applications. Based on Ceragon s field-proven 1500HP technology, RFU-HS supports capacities of up to 500 Mbps for TDM and IP interfaces. With its high transmit power, FibeAir RFU-HS is designed to enable high quality wireless communication in the most cost-effective manner, reaching over longer distances while enabling the use of smaller antennas Main Features of RFU-HS Frequency range Operates in the frequency range of 6-8 GHz Ultra high transmit power - Up to 30 dbm for longer distances, enhanced availability Configurable Modulation QPSK 256 QAM Configurable Channel Bandwidth 3.5 MHz 56MHz Direct or remote mount - Flexible installation saves costs and reduces transmission loss Supported configurations: direct and remote mount direct and remote mount direct and remote mount remote mount XPIC and CCDP Built-in XPIC (Cross Polarization Interference Canceller) and Co-Channel Dual Polarization (CCDP) ATPC (Automatic Tx Power Control) Simple and Easy Installation Ceragon Proprietary and Confidential Page 160 of 225

161 6.5.2 RFU-HS Frequency Bands Frequency Band Frequency Range (GHz) Channel Bandwidth Standard L6 GHz to /56MHz ITU-R F.383 U6 GHz to MHz to 40/56 /60 MHz ITU-R F to MHz to 28/56 MHz ITU-R F.385 Annex 4 7 GHz to /56 MHz ITU-R F.385 Annex to /56 MHz ITU-R F.385 Annex to MHz ITU-R F.386 Annex 1 8 GHz to MHz to 28/56 MHz ITU-R F.386 Annex to MHz to 28/56 MHz ITU-R F.386 Annex 4 Ceragon Proprietary and Confidential Page 161 of 225

162 6.5.3 RFU-HS Mechanical, Electrical, and Environmental Specifications RFU-HS Mechanical, Electrical, and Environmental Specifications RFU Dimensions Height: 409mm Width: 286 mm Depth: 86 mm Weight: 8 kg RFU Antenna Connection IDU-RFU Connection Standard flexible waveguide (frequency dependent) Coaxial cable RG-223 (100 m/300 ft), Belden 9914/RG-8 (300 m/1000 ft) or equivalent, N-type connectors (male) Maximum System Power Consumption (IDU and RFU) 1+0: 88W 1+1: 134W Storage Transportation Operating Temperature ETS class T1.2, with a temperature range of -25 C to+85 C. ETS class 2.3, with a temperature range of -40 C to+85 C. Temperature range for continuous operating temperature with high reliability: -33 C to +55 C (-27 F to 131 F) Temperature range for exceptional temperatures; tested successfully, with limited margins: -45 C to +60 C (-49 F to 140 F) Relative Humidity 5% to 100% Power Supply to -72 VDC (up to -57 VDC for USA market) RFU-HS Antenna Types The following antennas support direct and remote mount installations for RFU-HS. Vendor Frequency Band Diameter Manufacturer PN Marketing Model Andrew 7/8 GHz 4ft VHLP4-7W-CR3 A-4-7_8-A Andrew 7/8 GHz 6ft VHLP6-7W-CR3 A-6-7_8-A RFS 6L 4ft SU4-59CVA A-4-6L-R RFS 6L 6ft SU6-59CVA A-6-6L-R RFS 6U 4ft SU4-65CVA A-4-6H-R Ceragon Proprietary and Confidential Page 162 of 225

163 Vendor Frequency Band Diameter Manufacturer PN Marketing Model RFS 6U 6ft SU6-65CVA A-6-6H-R RFS 7/8 GHz 4ft SB4-W71CVA A-4-7_8-R RFS 7/8 GHz 6ft SU6B-W71CVA A-6-7_8-R Xian Putian 6L 4ft WTG12-58DAR A-4-6L-X Xian Putian 6L 6ft WTG18-58DAR A-6-6L-X Xian Putian 6U 4ft WTG12-64DAR A-4-6H-X Xian Putian 6U 6ft WTG18-64DAR A-6-6H-X Xian Putian 7/8 GHz 4ft WTG12-W71DAR A-4-7_8-X Xian Putian 7/8 GHz 6ft WTG18-W71DAR A-6-7_8-X RFU-HS Antenna Connection The RFU is connected to the antenna via a flexible waveguide (which is frequency-dependent), in accordance with the following table. (The antenna type and the waveguide flanges are imperial.) Frequency (GHz) Waveguide Standard Waveguide Flange 6L WR137 CPR137F 6H WR137 CPR137F 7 WR112 CPR112F 8 WR112 CPR112F RFU-HS Mediation Device Losses The following table lists branching losses for RFU-HS antennas. Configuration Interfaces 6-8 GHz Flex WG Remote Mount antenna Added on remote mount configurations Integrated antenna Integrated antenna HSB Main TR 1.6 Integrated antenna with asymmetrical coupler Secondary TR /2+2 HSB Main TR 1.6 Remote antenna with asymmetrical coupler Secondary TR SP (with CPLR) Integrated antenna DP Remote mount antenna 4 Ceragon Proprietary and Confidential Page 163 of 225

164 For RFU-HS transmit power specifications: RFU-HS Transmit Power (dbm) For RFU-HS receiver threshold specifications: RFU-HS Receiver Threshold (RSL) BER = 10-6) Ceragon Proprietary and Confidential Page 164 of 225

165 6.6 RFU-SP FibeAir RFU-SP supports multiple capacities, frequencies, modulation schemes, and configurations for various network requirements. RFU-SP operates in the frequency range of 6-8 GHz, and supports capacities of 40 Mbps to 400 Mbps for TDM and IP interfaces. The capacity can easily be doubled using XPIC Main Features of RFU-SP Frequency Range Operates in the frequency range of 6-8 GHz. Configurable Capacity from 40 Mbps to 500 Mbps. Configurable Modulation QPSK 256 QAM Configurable Channel Bandwidth 3.5 MHz 56MHz Antenna Mount Direct or remote. Main Configurations 1+1, 1+0, 2+0 XPIC and CCDP Built-in XPIC and Co-Channel Dual Polarization (CCDP) ATPC (Automatic Tx Power Control) Simple and Easy Installation Ceragon Proprietary and Confidential Page 165 of 225

166 6.6.2 RFU-SP Frequency Bands The frequency band of each radio is listed in the following table. RFU-SP Frequency Bands Frequency Band Frequency Range (GHz) Channel Bandwidth L6 GHz to /56MHz U6 GHz to MHz to 40/56 /60 MHz to MHz to 28/56 MHz 7 GHz to /56 MHz to /56 MHz to MHz 8 GHz to MHz to 28/56 MHz to MHz to 28/56 MHz Ceragon Proprietary and Confidential Page 166 of 225

167 6.6.3 RFU-SP Mechanical, Electrical, and Environmental Specifications RFU-SP Mechanical, Electrical, and Environmental Specifications RFU Dimensions Height: 409mm Width: 286 mm Depth: 86 mm Weight: 8 kg RFU Antenna Connection IDU-RFU Connection Standard flexible waveguide (frequency dependent) Coaxial cable RG-223 (100 m/300 ft), Belden 9914/RG-8 (300 m/1000 ft) or equivalent, N-type connectors (male) Maximum System Power Consumption (IDU and RFU) 1+0: 88W 1+1: 130W Storage Transportation Operating Temperature ETS class T1.2, with a temperature range of -25 C to+85 C. ETS class 2.3, with a temperature range of -40 C to+85 C. Temperature range for continuous operating temperature with high reliability: -33 C to +55 C (-27 F to 131 F) Temperature range for exceptional temperatures; tested successfully, with limited margins: -45 C to +60 C (-49 F to 140 F) Relative Humidity 5% to 100% Power Supply to -72 VDC (up to -57 VDC for USA market) Ceragon Proprietary and Confidential Page 167 of 225

168 6.6.4 RFU-SP Direct Mount Installation The following antennas support direct and remote mount installations: RFU-HS-SP Antennas Vendor Frequency Band Diameter Manufacturer PN Marketing Model Andrew 7/8 GHz 4ft VHLP4-7W-CR3 A-4-7_8-A Andrew 7/8 GHz 6ft VHLP6-7W-CR3 A-6-7_8-A RFS 6L 4ft SU4-59CVA A-4-6L-R RFS 6L 6ft SU6-59CVA A-6-6L-R RFS 6U 4ft SU4-65CVA A-4-6H-R RFS 6U 6ft SU6-65CVA A-6-6H-R RFS 7/8 GHz 4ft SB4-W71CVA A-4-7_8-R RFS 7/8 GHz 6ft SU6B-W71CVA A-6-7_8-R Xian Putian 6L 4ft WTG12-58DAR A-4-6L-X Xian Putian 6L 6ft WTG18-58DAR A-6-6L-X Xian Putian 6U 4ft WTG12-64DAR A-4-6H-X Xian Putian 6U 6ft WTG18-64DAR A-6-6H-X Xian Putian 7/8 GHz 4ft WTG12-W71DAR A-4-7_8-X Xian Putian 7/8 GHz 6ft WTG18-W71DAR A-6-7_8-X RFU-SP Antenna Connection RFU-SP is connected to the antenna via a flexible waveguide, which is frequency-dependent, in accordance with the following table. Frequency (GHz) Waveguide Standard Waveguide Flange 6L WR137 CPR137F 6H WR137 CPR137F 7 WR112 CPR112F 8 WR112 CPR112F Ceragon Proprietary and Confidential Page 168 of 225

169 6.6.6 RFU-SP Mediation Device Losses The following table lists branching losses for RFU-SP antennas. Configuration Interfaces 6-8 GHz Flex WG Remote Mount antenna Added on remote mount configurations Integrated antenna Integrated antenna HSB Main TR 1.6 Integrated antenna with asymmetrical coupler Secondary TR /2+2 HSB Main TR 1.6 Remote antenna with asymmetrical coupler Secondary TR SP (with CPLR) Integrated antenna DP Remote mount antenna 4 For RFU-SP transmit power specifications: RFU-SP Transmit Power (dbm) For RFU-SP receiver threshold specifications: RFU-SP Receiver Threshold (RSL) BER = 10-6) Ceragon Proprietary and Confidential Page 169 of 225

170 P P Mechanical, Electrical, and Environmental Specifications 1500P Mechanical, Electrical, and Environmental Specifications RFU Dimensions IDU-RFU Connection Diameter: 270 mm (10.8 ) Depth: 140 mm (4.5 ) Weight: 8 kg (18 lbs) Coaxial cable RG-223 (100 m/300 ft), Belden 9914/RG-8 (300 m/1000 ft) or equivalent, N-type connectors (male) Maximum System Power Consumption (IDU and RFU) 1+0: 65W 1+1: 105W Storage Transportation Operating Temperature ETS class T1.2, with a temperature range of -25 C to+85 C. ETS class 2.3, with a temperature range of -40 C to+85 C. Temperature range for continuous operating temperature with high reliability: -33 C to +55 C (-27 F to 131 F) Temperature range for exceptional temperatures; tested successfully, with limited margins: -45 C to +60 C (-49 F to 140 F) Relative Humidity 5% to 100% Power Supply to -72 VDC (up to -57 VDC for USA market) Ceragon Proprietary and Confidential Page 170 of 225

171 P Mediation Device Losses The following table lists branching losses for 1500P antennas. 1500P Mediation Device Losses Configuration Interfaces 11 GHz GHz GHz GHz Flex WG Remote Mount antenna Added on remote mount configurations Integrated antenna Integrated antenna HSB Main TR Integrated antenna with asymmetrical coupler Secondary TR /2+2 HSB Main TR Remote antenna with asymmetrical coupler Secondary TR For 1500P transmit power specifications: 1500P Transmit Power (dbm) For 1500P receiver threshold specifications: 1500P Receiver Threshold (RSL) BER = 10-6) Ceragon Proprietary and Confidential Page 171 of 225

172 7. Typical Configurations This chapter includes: Single Pipe Configurations Multiple Pipe Configurations (Chain/Node Sites) Ceragon Proprietary and Confidential Page 172 of 225

173 7.1 Single Pipe Configurations Supported Configurations Single Pipe IP-10Q single pipe configurations provide flexible Line protection, completely independent from radio configuration Multi-Radio (XPIC optional) 1+1 HSB (optional SD with BBS) Seamless Upgradeability An IP-10Q system is scalable. Using the same chassis, you can upgrade your system s capacity by adding additional carriers. Seamless Upgradeability Single Pipe Ceragon Proprietary and Confidential Page 173 of 225

174 Configuration FibeAir IP-10Q Typical Configurations 1+0 Ceragon Proprietary and Confidential Page 174 of 225

175 Multi-Radio Configuration FibeAir IP-10Q Typical Configurations 2+0 Multi-Radio Ceragon Proprietary and Confidential Page 175 of 225

176 7.2 Multiple Pipe Configurations (Chain/Node Sites) Seamless Upgradeability Multiple Pipes A multiple pipe configuration is scalable from uni-directional to multidirectional. Seamless Upgradeability Multiple Pipes- 1+0 Seamless Upgradeability Multiple Pipes- 2+0 and 2 x 2+0 Ceragon Proprietary and Confidential Page 176 of 225

177 Up to 2 x 2+0 Multi-Radio (XPIC Optional) and 1+1 HSB (BBS Space Diversity Optional) Ceragon Proprietary and Confidential Page 177 of 225

178 Multi-Radio East/West Configuration 2+0 Multi-Radio East/West Configuration Ceragon Proprietary and Confidential Page 178 of 225

179 8. FibeAir IP-10Q Management This chapter includes: Management Overview Management Communication Channels and Protocols Web-Based Element Management System (Web EMS) Command Line Interface (CLI) Floating IP Address In-Band Management Out-of-Band Management System Security Features Ethernet Statistics Software Update Timer CeraBuild Ceragon Proprietary and Confidential Page 179 of 225

180 8.1 Management Overview The Ceragon management solution is built on several layers of management: NEL Network Element-level CLI EMS HTTP web-based EMS NMS and SML NetMaster or PolyView platform Each IP-10 Network Element includes an HTTP web-based element manager (CeraWeb) that enables the operator to perform element configuration, RF, Ethernet, and PDH performance monitoring, remote diagnostics, alarm reports, and more. In addition, Ceragon provides an SNMP V1/V2c/V3 northbound interface on the IP-10Q. Ceragon s management suite also includes a number of CeraBuild tools, which ease the operator s task of installing, maintaining, and provisioning Ceragon equipment. Ceragon offers NetMaster and PolyView network management systems (NMS). Both NetMaster and PolyView provide centralized operation and maintenance capability for the complete range of network elements in an IP-10Q system. In addition, management, configuration, and maintenance tasks can be performed directly via the IP-10Q Command Line Interface (CLI). Integrated IP-10Q Management Tools Ceragon Proprietary and Confidential Page 180 of 225

181 8.2 Management Communication Channels and Protocols Related Topics: Secure Communication Channels Network Elements can be accessed locally via serial or Ethernet management interfaces, or remotely through the standard Ethernet LAN. The application layer is indifferent to the access channel used. PolyView can be accessed through its GUI interface application, which may run locally or in a separate platform; it also has an SNMP-based northbound interface to communicate with other management systems. Dedicated Management Ports Port number Protocol Packet structure Details 161 SNMP UDP Sends SNMP Requests to the network elements 162 Configurable SNMP (traps) UDP Sends SNMP traps forwarding (optional) 25 SMTP (mail) TCP Sends PolyView reports and triggers by (optional) 69 TFTP UDP Uploads/ downloads configuration files (optional) 80 HTTP TCP Manages devices 443 HTTPS TCP Manages devices (optional) From 21 port to any remote port (>1023) FTP Control Port TCP Downloads software and configuration files. (FTP Server responds to client's control port) (optional) From Any port (>1023) to any remote port (>1023) FTP Data Port TCP Downloads software and configuration files. The FTP server sends ACKs (and data) to client's data port. Optional FTP server random port range can be limited according to need (i.e., according to the number of parallel configuration uploads). All remote system management is carried out through standard IP communications. Each NE behaves as a host with a single IP address. The communications protocol used depends on the management channel being accessed. Ceragon Proprietary and Confidential Page 181 of 225

182 As a baseline, these are the protocols in use: Standard HTTP for web-based management Standard telnet for CLI-based management PolyView uses a number of ports and protocols for different functions: PolyView Server Receiving Data Ports Port number Protocol Packet structure Details 162 Configurable 4001 Configurable SNMP (traps) UDP Receive SNMP traps from network elements Propriety TCP CeraMap Server 69 TFTP UDP Downloads software and files (optional) 21 FTP Control Port TCP Downloads software and configuration files. (FTP client initiates a connection) (optional) To any port (>1023) from any Port (>1023) 9205 Configurable 9207 Configurable FTP Data Port TCP Downloads software and configuration files.(ftp Client initiates data connection to random port specified by server) (optional) FTP Server random port range can be limited according to needed configuration (number of parallel configuration uploads). Propriety TCP User Actions Logger server (optional) Propriety TCP CeraView Proxy (optional) Web Sending Data Ports Port number Protocol Packet structure Details 80 HTTP TCP Manages device 443 HTTPS TCP Manages device (optional) Web Receiving Data Ports Port number Protocol Packet structure Details 21 FTP TCP Downloads software files (optional) Data port FTP TCP Downloads software files (optional) Additional Management Ports for IP-10Q Port number Protocol Packet structure Details 23 telnet TCP Remote CLI access (optional) 22 SSH TCP Secure remote CLI access (optional) Ceragon Proprietary and Confidential Page 182 of 225

183 8.3 Web-Based Element Management System (Web EMS) The CeraWeb Element Management System (Web EMS) is an HTTP web-based element manager that enables the operator to perform configuration operations and obtain statistical and performance information related to the system, including: Configuration Management Enables you to view and define configuration data for the IP-10Q system. Fault Monitoring Enables you to view active alarms. Performance Monitoring Enables you to view and clear performance monitoring values and counters. Maintenance Association Identifiers Enables you to define Maintenance Association Identifiers (MAID) for CFR protection. Diagnostics and Maintenance Enables you to define and perform loopback tests, software updates, and IDU-RFU interface monitoring. Security Configuration Enables you to configure IP-10Q security features. User Management Enables you to define users and user groups. A Web-Based EMS connection to the IP-10Q can be opened using an HTTP Browser (Explorer or Mozilla Firefox). The Web EMS uses a graphical interface. All system configurations and statuses are available via the Web EMS, including all L2-Switch configurations such as port type, VLANs, QoS. The Web EMS shows the actual chassis configuration and provides easy access to any IDU in the chassis. Ceragon Proprietary and Confidential Page 183 of 225

184 8.4 Command Line Interface (CLI) A CLI connection to the IP-10Q can be opened via terminal (serial COM, speed: , Data: 8 bits, Stop: 1 bit, Flow-Control: None), or via telnet (SSH is supported as well). The Terminal format should be VT-100 with a screen definition of 80 columns X 24 rows. All parameter configurations can be performed via CLI. All IDUs in a chassis can be accessed by the CLI interface, by using a command which enables the user to login to any slot in the chassis Text CLI Configuration Scripts CLI configuration text scripts, written in Ceragon CLI format, can be downloaded into the IDU. It is not possible to upload the IDU s configuration into a text file. CLI scripts can only be downloaded and handled via CLI. CLI scripts cannot be downloaded via the Web EMS. The user can perform the following operations on CLI scripts: Set the file name of the script: Download CLI script file to the IDU Download the CLI script file: Get the status of the downloaded script. Show the last downloaded CLI script content. Execute (activate) a CLI script. Delete the current script which resides inside the IDU. Protection copy-to-mate command Ceragon Proprietary and Confidential Page 184 of 225

185 8.5 Floating IP Address The floating IP address feature provides a single IP address that will always provide direct access to the currently active main unit in a 1+1 HSB configuration. This is used primarily for web-based management and telnet access. The user can configure a floating IP address in the active unit, and this IP address will be automatically copied to the standby unit. The following limitations apply: The floating IP address must be different from the system IP address. The floating IP address must be in the same subnet as the system IP address. The remote floating IP address can be viewed and configured using the localremote channel. The individual units IP addresses are maintained in order to provide a mechanism to connect directly to the standby unit should this be necessary for any reason. For SNMP access, a mechanism exists to similarly enable automatic access to active protected extension units. Note that when using the SNMP protocol, the actual IDU being accessed depends on the community/context string. The floating IP address feature can still be used to ensure access if one of the main units fails. The floating IP mechanism can be enabled or disabled. When it is enabled, then upon a protection switch, the existing floating IP address is assigned to the unit that was previously in standby mode and has switched to active mode. This unit will have a different MAC address than the previously active IDU. For this reason, a gratuitous ARP (GARP) message is automatically sent after the switch. However, when connected directly to some older network equipment, reestablishment of the management Ethernet ports link may take a few seconds after a protection switch. In this case, the GARP message may be lost. For this reason, users can configure a number of GARP transmission retries (default is 5 retries, maximum is 10). Retries will be sent one time per second. In the unlikely case of repeated protection switches (which may take place as a result of permanent radio channel problems), communication may be lost due to the fact that the ARP changes are taking place once every few seconds. In this case, the floating IP address will be automatically locked to one of the IDUs so that users can maintain remote management access to the system. Note that the IDU may be a standby unit. The IP address will automatically return to the active unit when the situation stabilizes. Alternatively, users can access any of the IDUs in the chassis using their local IP addresses. Ceragon Proprietary and Confidential Page 185 of 225

186 8.6 In-Band Management FibeAir IP-10Q can optionally be managed In-Band, via its radio and Ethernet interfaces. This method of management eliminates the need for a dedicated interface and network. In-band management uses a dedicated management VLAN, which is user-configurable. With In-Band management, the remote IDU is managed by specific frames that are sent as part of the traffic. These frames are identified as management frames by a special VLAN ID configured by the user. This VLAN ID must be used only for management. It is not possible to configure more than a single VLAN ID for management. Note: It is strongly recommended to classify the management VLAN ID to the highest queue, in order to ensure the ability to manage remote units even under congestion scenarios. The local unit is the gateway for In-Band management. The remote unit is managed via its traffic ports (the radio port, for example), so that no management ports are needed In-Band Management Isolation This feature is designed for operators that provide Ethernet leased lines to third party users. The third party user connects its equipment to the Ethernet interface of the IP-10Q, while all the other network interfaces, particularly the radios, are managed by the carrier of carriers user. In that case, management frames that are sent throughout the network to manage the carrier of carrier equipment must not egress the line interfaces that are used by the third party customer, since these frames will, in effect, spam the third party user network. The following figure describes the management blocking scenario. In-Band Management Isolation 3 rd Party User Network IP-10 Block provider s management Frames Carrier of carriers network (Provider Network) Mng Frames Mng Frames IP-10 Block provider s management Frames 3 rd Party User Network Provider Network Management Center Ceragon Proprietary and Confidential Page 186 of 225

187 8.7 Out-of-Band Management With Out-of-Band management, the remote system is managed using an Ethernet management channel provided by a third party equipment. Eth2 and Eth3 can be used to chain management from one shelf to another. Management ports can also be enabled on the extension slots (slot 2, slot 3 and slot 4). These management ports on the extension slots will have the same connectivity as management ports on slot 1. Important: To avoid management loops, do not connect two management ports of the same chassis to an external switch at the same time. Management frames that ingress from the management ports must not be VLAN tagged. Tagged frames will be discarded. In HSB protection mode, Out-of-Band management ports must be protected by a split Ethernet cable. In this configuration, management ports must be configured as FE interfaces. Ceragon Proprietary and Confidential Page 187 of 225

188 8.8 System Security Features To guarantee proper performance and availability of a network as well as the data integrity of the traffic, it is imperative to protect it from all potential threats, both internal (misuse by operators and administrators) and external (attacks originating outside the network). System security is based on making attacks difficult (in the sense that the effort required to carry them out is not worth the possible gain) by putting technical and operational barriers in every layer along the way, from the access outside the network, through the authentication process, up to every data link in the network Ceragon s Layered Security Concept Each layer protects against one or more threats. However, it is the combination of them that provides adequate protection to the network. In most cases, no single layer protection provides a complete solution to threats. The layered security concept is presented in the following figure. Each layer presents the security features and the threats addressed by it. Unless stated otherwise, requirements refer to both network elements and the NMS. Security Solution Architecture Concept Ceragon Proprietary and Confidential Page 188 of 225