You are on page 1of 17

White Paper

Eclipse in the Mobile Network ETSI

Base stations that can connect devices at up to 50 Mbit/s or more will be operational within
the next few years. For the backhaul network the extra capacity needed and the mix of
services will require changes in the technology used – IP/Ethernet will become the transport
technology of choice.
This paper introduces the technologies for next generation backhaul networks, and the
connection and management solutions provided by Eclipse Carrier Ethernet over Wireless
platforms from Harris Stratex Networks.

Data services will grow quickly to use more network capacity than voice. More network
capacity translates to more backhaul capacity. Coupled with this is the recognition that
Eclipse solutions Ethernet is the transport media of choice for expanded backhaul services.
are optimized for
For many operators the introduction of Ethernet will be on the back of existing TDM network
now and next connections given their huge investment in its infrastructure. This will typically involve gradual
generation mobile migration using data overlay, with a decision at some future point to change to an all packet-
based network. Other operators may elect to forgo migration and completely replace existing
backhaul TDM networks using Ethernet. Pseudowires will be used to support legacy TDM connections.
networks. Whatever the direction, Eclipse provides optimized wireless backhaul solutions through its
unique packet and circuit switched architecture.
The value-add
• The extended packet plane supports multiple GigE connections to 1.4 Gbit/s. Link
node concept capacities can be configured to 360 Mbit/s, 720 Mbit/s CCDP, or 1.4 Gbit/s
provides CCDP/Quattro.
maximum • The Liquid Bandwidth circuit plane supports native mixed mode operation with Super
performance with PDH capacities to 100xE1 and Ethernet to 200 Mbit/s. Services are split between
TDM and Ethernet in 1xE1 / 2 Mbit/s steps to the maximums, to accommodate a low-
lowest cost and risk PDH now, and Ethernet tomorrow transport philosophy.
risk. Assisting the upgrade route are Eclipse options for better spectrum efficiency. More capacity
can be transported on existing channels using high-order modulation, adaptive modulation or
co-channel operation.
Similarly, carrier grade Ethernet performance is made possible via an intelligent layer 2
switch to ensure that the transport of Ethernet data is no less secure than for TDM. When
coupled with advanced traffic prioritization, RWPR, link aggregation, pseudowires, network
synchronization, MPLS, bandwidth optimization and traffic aggregation, there is a Harris
Stratex solution for all network topologies.
The backbone for this capability is the Eclipse INU where plug-in modules provide uniquely
flexible and scalable platforms for lowest incremental cost and maximum value-add. Where
the plug-in flexibility of the INUs is not required, such as at edge sites, Eclipse IDUs provide
Headquarters cost and performance optimized solutions for Ethernet with or without PDH. In most situations
Harris Stratex Networks, Inc. Eclipse eliminates the need for expensive external network devices.
Research Triangle Park
637 Davis Drive
Finally, Eclipse comes with an assurance from Harris Stratex that value-adds will continue to
become available to existing and new Eclipse customers to deliver more features and more
performance. It is a promise of a low-risk and future-proof investment in Eclipse.
North Carolina 27560
United States
Tel: 919-767-3230
Fax: 919-767-3233 7/25/2008 EclipseInTheMobileNetwork_ETSIv2.doc Page 1 of 17

especially so where the backhaul networks are. White Paper Introduction Typically a small number of E1s has been sufficient to service 2G and 2.5G base stations. The mobile data Going forward. and not just one operator. but with the data capacity needed for advanced 3G and 4G HSPA/LTE applications. do you overlay with Ethernet to deliver the extra capacity. which means operator revenues will be uncoupled from the traditionally linear returns on provisioning for voice growth. new technologies and strategies are required. This projection assumes that the cost to access mobile data closely aligns with subscriber expectations for accessing data via wired or WiFi connections. where their per-hop behaviour massive changes capabilities can provide new efficiencies in the transport of different types of traffic . and a programmable total nodal throughput of over 1 Gbit/s Ethernet and up to 100xE1 or 2xSTM1. the most cost effective backhaul technology to deliver more capacity. For 4G/LTE. This need for more capacity must be provided more intelligently. capabilities and In networks where wireless provides the backhaul. and more efficiently. Of special note is that one Eclipse node now supports up to six links. is Carrier Ethernet. or will be required to support multiple services and customers – not just cellular mobile. . Figure 1.for in mobile network different customers. It is also the ideal precipitating tecnology for use in MPLS or PBB-TE converged networks. the cost of mobile data to subscribers must be lower than for voice for the same data bandwidth. there is a cost efficient solution using Eclipse. Some features are subject to availability.doc Page 2 of 17 Copyright © 2008 Harris Stratex Networks. you simply increase the capacity of current TDM links. with some city sites as high as 50 Mbit/s. There is also a need to ensure service continuity for old and new technologies. Certainly. how do you upgrade wireless connections to deliver more capacity? Do infrastructure. the 3G data revolution is expected to require from 10xE1 or 20 Mbit/s within 3 years. The Uncoupling of Voice and Data Revenues This need for more capacity brings pressure on the backhaul network. all rights reserved. where instead of current 2xE1 / 4 Mbit/s connections per 2G BTS. or do you move to an all-Ethernet solution? Whichever way is forward. more evolution is intelligently. More Capacity The expected global growth of mobile broadband data services through EDGE/HSPA/LTE evolution will see data traffic exceeding voice within a relatively short period of time. given that operators will want to maximize investments in existing 2G/3G infrastructure. For example. these developments raise a number of issues. flexibility and QoS features needed to provide a complete solution – from the core MSC to the base stations. 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. figures as high as 100 Mbit/s have been forecast. Hence operators must implement more cost-efficient solutions for delivering more network capacity. It provides the scalability.

Ultimately. .000s of individual Carrier Ethernet services over local. suppliers. • Flexibility: Ethernet supports speeds from 1 to 10 Gbit/s in 1 Mbit/s steps. and quality of service (QoS). scalability. traffic aggregation. Some.doc Page 3 of 17 Copyright © 2008 Harris Stratex Networks. However. Ethernet also supports easy convergence of mobile backhaul with other network applications on in-house or third-party networks. flexibility. it’s extremely wide usage means it is better supported by network operators and the manufacturing and support industries. an all Ethernet backhaul is considered a must. easy scalability. with the roll-out of LTE. giving operators the choice of one or both technologies. pseudowires and network synchronization are key considerations in migrating from TDM to Ethernet. More Intelligence We have mentioned the Ethernet benefits of cost. • Scalability: Ethernet readily lends itself to servicing many 100. advantages of cost. pseudowires. high priority voice data can be given right of way over lower priority and non-real-time for next data services. Mixed Mode Mixed mode is about side-by-side transport of TDM and native Ethernet data. in most networks it is expected that the rollout of Ethernet-capable base stations will complement an installed base of TDM- only base stations. Ethernet versus TDM It is as well to mention some general advantages of packet-based IP/Ethernet over that of circuit-switched TDM transport at this point. Ethernet delivers more cost-effective bandwidth than other technologies. White Paper Base Station Readiness Base stations are now becoming available with TDM and Ethernet access interfaces. all rights reserved. Others like MPLS and PBB-TE are particularly relevant to efficient and robust transportation in a converged network. And with carrier-grade performance now on offer from some networks. meaning that for the medium term at least. national and international connections. If bottlenecks occur. where Ethernet is used to meet the rapidly growing data demand. reliability and availability of Ethernet at least matches that on offer from traditional TDM transport technologies. like mixed-mode. the overlaying of a TDM network with Ethernet. metro. Some features are subject to availability. 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. there will be a mix of TDM and Ethernet in the backhaul. • Cost: As the protocol of choice for Internet and business-based intranets. generation The bottom line is that Ethernet is considered the way forward for a data-driven expansion of backhaul the mobile backhaul network. flexibility and QoS. bandwidth optimization. and IP based OAM for end-to-end traffic and performance management.gone are the needs to consider PDH connections to an expensive SDH core. is the technology • QoS: Ethernet supports operator-friendly prioritization of traffic. MPLS / PBB-TE. such as mixed mode. In this section we introduce technologies that add intelligence to the way capacity is provided and managed in the network. In a backhaul network it means Ethernet can provide an end-to-end solution from the BTS to the network core .

Currently IEEE 1588v2 and synchronous Ethernet have the front running. which operate as virtual tunnels across provider network(s) to support legacy traffic. But pseudowires do impose an overhead – typically an additional 10 to 20% % is needed over and above the native TDM bandwidth. Pseudowires do not support a robust solution for transporting the timing signals needed for base station synchronization. The More capacity one proviso is that such a solution must be cost-efficient. it must represent a lower cost than must be provided a switch to all-Ethernet at the outset. Industry feedback indicates that non-proprietary solutions are needed. but at the same time support efficient migration to an all- Ethernet network when needed. IEEE 1588v2. Figure 3. Options include synchronous Ethernet (G. but does have some 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. Some features are subject to availability. White Paper Such a strategy has merit from the viewpoint of maximizing the use of existing TDM infrastructure.doc Page 4 of 17 Copyright © 2008 Harris Stratex Networks. This is where pseudowires provide an answer . particularly so in converged networks where standardized operating practices and technologies are key requirements. and proprietary pseudowire adaptive timing solutions. more intelligently Figure 2. Synchronization is also an issue. Pseudowires Synchronization An IP/MPLS or all-Ethernet network should support transmission of the frequency and phase synchronization requirements for base stations to the standards expected for 4G implementations. NTP.8262). . All-Ethernet and Pseudowires Replacing existing backhaul infrastructure with an all Ethernet solution will invariably require accommodation of existing TDM network connections.existing TDM services are transported end-to-end in the network on pseudowires. all rights reserved. while minimizing the risk associated with introducing a new technology. Mixed Mode to maximize ROI and minimize disruption to existing services. IEEE 1588v2 meets requirements for precision frequency and phase synchronization. GPS timing has also been promoted as a solution.

and mpt-to-mpt (any-to-any) Ethernet Virtual Private LAN Service (VPLS). though all-MPLS or all PBB-TE solutions will have their place depending on technologies. all rights reserved. SDH and PDH.doc Page 5 of 17 Copyright © 2008 Harris Stratex Networks. White Paper traffic loading issues. . improved resiliency with pre-defined failover scenarios. identified by a label placed in each packet. 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. make better use of network capacity with multiple accommodation for a wide range of service types and bandwidth plans. ATM. They also provide networks. Essentially. it is particularly suited to networks that carry different mixtures of traffic over different network connections for multiple users. congestion and differentiated (prioritized) services. It works with IP to support both layer 3 virtual private network offers networks (VPNs) and layer 2 pseudowires. Synchronous Ethernet does not provide phase synchronization.MPLS Transport MPLS (T-MPLS) is an emerging subset of MPLS. for example. synchronous Ethernet in the core and IEEE 1588v2 in the Metro and access networks. Figure 4. and operator preferences. It operates with IP. with PBB-TE in Metro and access networks. They are designed to speed up network traffic flow. and is directed (tunneled) through the MPLS network based on this label. As such. IP/MPLS Virtual Private LAN Service T. traffic enters and exits an MPLS network via Provider Edge (PE) switch/routers. services onto one MPLS ‘converged’ MPLS is a mature ITU / IETF standard. Ultimately the two may co-exist as merged solutions with. pt to significant mpt. At the PE egress point the MPLS label is stripped. It operates on similar lines to PBB-TE. Ingressing traffic is uniquely labeled (tagged) based on the desired destination and quality of service. and Frame Relay network-layer scale. but does provide frequency synchronization independent of traffic loading. protocols and over transport layers that include Ethernet. MPLS and PBB-TE The merging of MPLS and PBB-TE are “traffic engineering” technologies for converged networks. Its flexibility includes support for pt-to-pt. and support superior management and control features. It is optimized for Metro Ethernet network applications to provide purely connection-oriented services for managed point-to- point connections. economies of Its operation involves setting up a specific path for a given sequence of packets. The label switches manage outages. customer groupings. Current consensus supports MPLS in core networks. Some features are subject to availability. Figure 4 illustrates Ethernet multipoint layer 2 VPN operation over an MPLS network using the VPLS function. Ethernet services are delivered transparently between customer LANs at sites A to E.

techniques can Benefits include improved latency. and protocol acceleration. Eclipse Wireless Backhaul Solutions This section introduces Eclipse and its solutions for more backhaul capacity. Non- and traffic value data is removed for each voice channel. performance monitoring. . which acts to streamline data communications at the transport layer. diagnostics. aggregated. and A. Relevant OAM standards for Carrier Ethernet networks are ITU recommendation Y-1731 for the services layer (UNI to UNI). over typical circuit usage. Techniques include data compression based on the type of content. It has a flatter (layer 2 only) structure compared to MPLS. Data optimization An example is the use of packet and bandwidth optimization for GSM A. and switched. and IEEE 802. Some features are subject to availability. • The spectrum efficiencies offered with adaptive modulation and co-channel operation are described. In particular. overall improvements in bandwidth utilization can be as high as 4:1. White Paper PBB-TE PBB-TE is an emerging standard for point-to-point connection-oriented services in Ethernet networks. such as in a Metro network. When combined with data optimization techniques.doc Page 6 of 17 Copyright © 2008 Harris Stratex Networks. Metro and access NGN technologies are introduced. more efficiently. data reduction that reduces the number of bits needed to be transmitted. more intelligently. and the intelligence and capacity efficiencies offered for Core. The Metro Ethernet Forum (MEF) has also developed relevant standards and recommendations. and the migration of capacity from TDM to carrier- grade Ethernet is explained. Operation. maintenance and configuration. For example separate 2G and/or 3G circuit-switched connections are converted to packet-based data and aggregated (multiplexed) using the traffic aggregation capabilities of a layer 2 Ethernet switch.1ag for the connectivity layer. all rights reserved. to provide cost optimized solutions where a large amount of connection-oriented traffic needs to be hubbed. it introduces the unique switch-plane and packet- plane capabilities of the Eclipse nodes that together optimize transport options for Ethernet and TDM. provide significant Traffic Aggregation improvements in Traffic aggregation combines two or more data streams onto a common stream. and through bandwidth packet statistical multiplexing achieves improvements in bandwidth utilization of about 50% utilization.bis BTS links. 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. Administration and Maintenance (OAM) OAM is about end-to-end network management capabilities for fault detection/recovery. • The Eclipse platforms are described.bis data is extracted and converted to a aggregation packet format to deliver overall bandwidth savings of up to 50%. Data Optimization Data optimization is about reducing or compressing data so that more data can be sent over a given bandwidth. and more efficient use of existing infrastructure to eliminate or put back the need to update a network to deliver more capacity.

all rights reserved. One Eclipse INU network node. however. PDH and SDH. The IDUs support single-link 1+0 or 1+1 operation. Solutions are provided for switched GigE Ethernet. For GigE applications the IDU GE 20x provides up to 360 Mbit/s Ethernet throughput with up to 20xE1 tribs. SDH is needed in the wireless backhaul. directly supports up to three links. This represents a Super-PDH major performance improvement over the 63xE1 maximum on an STM1 (SDH) link. Ethernet + PDH. • If. Depending on the capacity and bandwidth selected. and circuit INUs for Nodes switched TDM on The INUs replace the traditional terminal or a common single-link based approach to networking with a nodal solution. migration from 5x Super PDH versus SDH to 100xE1 using In most wireless networks Eclipse Super PDH eliminates the need to have an SDH core. and minimizes costs. They are optimized for edge linking from an INU/INUe. They With packet comprise the node-based INU/INUe. not just point-to-point links. performance cabling and rack space. Super PDH provides a complete network solution for star and ring topologies all the way to the core. The same by traditional PDH hardware . White Paper Eclipse Platform Eclipse platforms are optimized for building networks.SDH offered more capacity more robustly than PDH. the INUe Eclipse INUs supports up to six links. and does so at a much lower cost migration from 10 when compared to an SDH+PDH solution. This nodal concept dramatically reduces equipment. to 200 Mbit/s. it refers to the ability to transport up to 100xE1 over a wireless link. and as the name suggests. 7/25/2008 EclipseInTheMobileNetwork_ETSIv2.doc Page 7 of 17 Copyright © 2008 Harris Stratex Networks. or where simple IDU-to-IDU linking is required. Eclipse also delivers with link options of STM1 and 2xSTM1. Radio paths and customer interfaces are customized by enable unique plug-in cards. • In a wireless backhaul network. the circuit switch SDH was generally adopted because of the limiting capacity and protection options offered plane. modulation rates range from QPSK to 256 QAM. Different models are available to support Ethernet. plane also However. PDH and SDH. delivers superior network flexibility and resilience. packet and circuit. to provide an unmatched range of options for capacity and channel efficiency. and terminal based IDUs. Some features are subject to availability. Eclipse Super PDH Super PDHTM operates on the Eclipse circuit switch-plane of the INUs. or the supports capacity typical 16xE1 or 20xE1 maximums of standard PDH links. . a single technology is the logical choice where it provides a complete solution – especially so when it also provides significant cost and performance benefits over other options. efficiencies on IDU Terminals wireless networks. now supported by two cost and switch planes. Eclipse Super PDH means that these reasons for installing an SDH core in a supports Ethernet backhaul network are no longer valid for wireless.

For example. the circuit and • Links will be expected to transport both TDM and Ethernet traffic natively. using plug-in cards. White Paper Eclipse Ethernet Eclipse links support native GigE or FastE Ethernet connections. • Links can be configured for PDH and Ethernet using Liquid Bandwidth. all rights reserved. • Links can be configured to 360 Mbit/s true Ethernet throughput. With both circuit and extended packet switch-planes. flexible and cost-efficient. RFC 2544 rated throughput figures. the extended packet plane capability is introduced Ethernet. at which point an TDM and operator can locally or remotely configure the capacity split between PDH and Ethernet in E1 Ethernet. the or 2 Mbit/s steps using the Liquid Bandwidth feature on the circuit switch-plane. . operators are expected to overlay with Ethernet to Ethernet is provide the extra capacity needed to support data-based services. and planes. • One Eclipse Node supports up to six links. with the burgeoning data traffic on an Ethernet overlay. not the layer 1 bit-rate figures used by some vendors to inflate their specs to 400 Mbit/s and beyond. It means Ethernet can be activated when and where needed in the network with minimum disruption.4 Gbit/s. Eclipse INU Packet and Circuit Switch Planes 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. Moreover. existing voice supported on both connections will be transported using legacy TDM links. Note that Harris Stratex uses layer 2. The ultimately to all-Ethernet. When coupled with Eclipse adaptive modulation. Figure 5. adding Ethernet to a Super-PDH link simply requires installation of a GigE or FastE plug-in. will be straight-forward. plane is specific to Then when more data capacity is needed. aggregation and CCDP capabilities. packet switch lowest cost and no risk. Some features are subject to availability. Figure 5 illustrates the packet and circuit switch plane architecture. with or without companion TDM traffic for mixed mode requirements. packet switch there will be an expectation that the migration from TDM-only to TDM + Ethernet. Eclipse plane supports INUs enable optimized native mixed mode solutions. with or without legacy PDH circuits. circuit switch This is what Eclipse is about. one Eclipse node supports multiple link connections with a combining capacity of more than 1.doc Page 8 of 17 Copyright © 2008 Harris Stratex Networks. • For many legacy PDH networks.

and modulation can also be selected from option menus. Eclipse Liquid Bandwidth Eclipse Wireless Packet Node The new Eclipse extended packet plane capability is unique . • The extended packet plane supports multiple GigE user interfaces with inter-card The packet switch Ethernet bridging and aggregation. to six 1+0 links. capacity. which indicates possible assignments to native Ethernet and to companion NxE1 capacity for a supports easy selected link capacity. protection. It means the native capacity capabilities on existing INUs is increased capacity of 360 five-fold. with each supporting native GigE to 360 Mbit/s.doc Page 9 of 17 Copyright © 2008 Harris Stratex Networks. which for many requirements will require nothing more than a configuration change using the Portal craft tool. White Paper Eclipse Liquid Bandwidth for Seamless Scalability Liquid BandwidthTM is the Eclipse ability to seamlessly assign link capacity to Ethernet. migration from Liquid Bandwidth means that the traffic over an Eclipse link can be user-assigned between PDH to native PDH and Ethernet – from PDH-only to PDH+Ethernet. Furthermore. It uses the Nx2 Mbit/s circuit switch plane to assign capacity in 2 Mbit/s / E1 steps to optimize Liquid bandwidth throughput granularity for network planning purposes. Some features are subject to availability. . Link Ethernet. plane supports up • The circuit plane supports the Liquid Bandwidth Ethernet and PDH connections. Figure 6. and to companion PDH traffic for native mixed mode delivers industry-leading flexibility and performance. the effectiveness of this gain can be magnified by the statistical Mbit/s. and aggregation of traffic. maximum air. channel bandwidth. This is illustrated in Figure 6. 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. and ultimately to all-Ethernet. all rights reserved. This capability is backward compatible with installed INUs to provide exceptional value-add on Eclipse networks. One Eclipse node now supports up to six separate links. multiplexing provided by traffic aggregation and data optimization options. Or up to five GigE links and one Liquid Bandwidth link for Ethernet and PDH. • One Eclipse Node is simply populated with the mix of cards needed at any time to each with a provide the required throughput.

High-order modulation is used to achieve maximum efficiency per channel. 204 Mbit/s or 100xE1 is supported on 40 MHz or 56 MHz channels using 128 QAM and 32 QAM respectively.5 to 56 MHz. Figure 7. Where more capacity is needed than can be provided over one link. 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. or 2xSTM1 in a 56 MHz channel. .STM1 is less spectrally efficient when transporting E1 streams.doc Page 10 of 17 Copyright © 2008 Harris Stratex Networks. all rights reserved. 9 illustrate some of the configurations made possible. 8. Similarly. For even higher capacities there are Eclipse 4+4 quattro solutions. 200xE1 or 4xSTM1. • 360 Mbit/s in a 56 MHz channel. The combined packet and circuit planes represent a breakthrough in wireless backhaul connectivity. Figure 9. White Paper Figures 7. Ethernet and PDH Aggregation Node One Eclipse INU or INUe supports multiple links. Note that 1xSTM1 only supports a maximum 63xE1 . Ring/Mesh Node connects support complex ring and aggregation network topologies. Some features are subject to availability. Examples include: • An industry-leading 190 Mbit/s or 93xE1 in a 28 MHz channel using the Liquid Bandwidth capability. • 1xSTM1 in a 28 MHz channel. Eclipse options include co-channel dual polarization (CCDP) to support 720 Mbit/s. Ring and Spur Node Link Capacity and Spectral Efficiency Eclipse link capacity options are achieved within standard ETSI channel bandwidths ranging from 3. and the advanced L2 switch and TDM cross- Figure 8.

• Link Aggregation. QoS and service MEF. • Harris Stratex Networks is a founding member of the MEF Mobile Backhaul Group. . Some features are subject to availability. or the DSCP bits in the Differentiated Services (DiffServ) field of an IP header. and best-quality Operation. It is especially relevant to wireless 7/25/2008 EclipseInTheMobileNetwork_ETSIv2.doc Page 11 of 17 Copyright © 2008 Harris Stratex Networks. all rights reserved. Such networks may require support for traffic routing. networks must be able to detect and recover from incidents without impacting users. Administration and Maintenance (OAM) in conjunction with the Harris Stratex ProvisionTM EMS. • RWPRTM (Resilient Wireless Packet Ring). Two or more links are combined into a single logical link. These include port and VLAN prioritization. • Q and Q-in-Q Tagging using the CoS/802. reliability. Tagging can be retained into an external network for downstream traffic management. aggregation and protection. reconvergence (service restoration) times are as low as 50 ms. and link aggregation: • Port Prioritization. Depending on the network topology. using features that include advanced QoS options. • Flow Control. whose aim is to promote and define the use of Carrier Ethernet services for mobile/cellular networks Carrier-grade Performance The technology required to deliver MEF 9 and MEF 14 compliance requires an intelligent Ethernet switch. VLAN Q and Q-in-Q tagging. outstanding reliability and scalability. Eclipse satisfies these requirements with features normally only found in advanced stand- alone switches. next generation • Eclipse GigE radios are certified compliant with the MEF 9 and MEF 14 backhaul Carrier Ethernet standards. • This certification provides an assurance that Eclipse GigE links will interoperate with other carrier Ethernet devices. White Paper Intelligent. scalability. RWPR enhances industry-standard RSTP (802. prioritization. fast- switched RSTP (RWPR) operation. with a traffic capacity that is the sum of the individual links. Advanced QoS and Performance Options MEF specifies that support must be provided for Service Level Agreements (SLAs) to deliver end-to-end performance matching over converged networks. there must be the ability to meet carrier- as defined by the grade requirements for standardized services. networks. Carrier Grade Ethernet For an Ethernet device to be considered carrier grade. Prioritizes traffic on one port over another. simple rule-based packet forwarding capabilities supported by basic L2 switches. technology for Eclipse Carrier Ethernet over Wireless has this capability.1p prioritization bits. In the Q-in-Q (backbone provider bridge) mode Eclipse can aggregate up to 4 separate customer VLANs onto a common radio channel. now and into the future. MEF 14 specifies the QoS (Quality of Service) parameters. MEF 9 specifies the UNI (Universal Network Interface). Provides a mechanism to throttle back data from sending devices to reduce demands on available Ethernet bandwidth. Ethernet traffic is prioritized on a frame-by-frame basis using the CoS (Class of Service) bits in the VLAN field of an Ethernet header. Additionally. Eclipse is compliant. is the management. As determined by the Metro Ethernet Forum (MEF). • VLAN (Tag) Prioritization. it must have intelligence beyond the Carrier Ethernet.1D-2004) with a unique rapid failure detection (RFD) capability to provide carrier- class network reconvergence times on Eclipse GigE ring and mesh networks.

performance monitoring. and performance monitoring at service (VLAN) and link levels. QoS priority settings are used to ensure all higher priority traffic continues to get through. .doc Page 12 of 17 Copyright © 2008 Harris Stratex Networks. Some features are subject to availability. management of the radio and Ethernet functions is fully integrated to services. MEF) call for network-wide fault detection/recovery. where Ethernet performance is being affected by radio performance. four for 1. Figure 10. IEEE 802. Eclipse offers Layer 2 and Layer 1 link aggregation options. Example ProVision EMS Screens for Network Health and Bandwidth Utilization 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. Two physical links are used for 720 Mbit/s.3ad) uses source and/or destination MAC address data in the Ethernet frame MAC/LLC header. and maintenance. all rights reserved. E-LAN OAM • With Eclipse. . Link aggregation also provides sub 50 ms redundancy. deliver maximum visibility to operators. ProVision additionally supports end-to-end network mapping. errors and discards. . For example.4 Gbit/s. the problem is easily diagnosed using common user-friendly interfaces. its traffic is redirected onto the remaining link. Uniquely.1ag. Advanced OAM for End-to-End Management OAM standards (ITU Y-1731. each is supported on its wide E-Line and own management system. Harris Stratex Eclipse and its ProVision EMS provide these capabilities end-to-end for the radio and L2 switch functions: ProVision EMS provides network. If one link fails. White Paper links when traffic capacities higher than the 300 to 360 Mbit/s maximums for a single link are required. Unlike L2 link aggregation it provides optimum payload sharing regardless of the throughput demands of individual user connections. Ethernet history. Relevant Ethernet and radio error events are supported by probable-cause and remedial advice. Ethernet diagnostics include RMON performance data. circuit provisioning. and Ethernet data- dashboards for throughput.Layer 2 link aggregation (802. making it ideal for router-router applications. • Where external switches are used in a wireless Ethernet network there is usually no management synergy between the switch and the radios. diagnostics.Layer 1 aggregation acts on the byte data stream. and to effect changes when needed. This means every device in the network must be visible to a network operator to provide the tools needed to determine device and network status and performance. or links. If the remaining link(s) do not have the capacity needed to avoid a traffic bottleneck.

Adaptive Modulation Illustration With Eclipse.5% of the time. it can be configured to ensure all high priority traffic continues to get through when path conditions deteriorate. It uses one of five automatically and dynamically switched modulations . For a given RF channel bandwidth of 7. This is where adaptive modulation provides a solution.999% availability under all path conditions.5% of the • By using less robust but more efficient modulation schemes. Less critical traffic is assigned to the higher modulations.999% availability. Most importantly. White Paper Adaptive Modulation In many instances spectrum availability is limited. Figure 11 illustrates the modulation/capacity steps and the percent availability over time. On a typical link this means higher capacity will be available for better than 99. Figure 11. QPSK. it dynamically changes the modulation so that the highest availability of capacity is provided at any given time. This is the purpose time. is increased when path conditions permit. for 99. adaptive modulation simply requires installation of a plug-in card. • Wireless links are traditionally engineered to carry traffic with a 99. • Adaptive modulation refers to the dynamic adjustment of modulation rate to ensure maximum data bandwidth is provided most of the time. of adaptive modulation. or 256 QAM . the available fade margin can be transformed into delivering more data throughput. a three- 7/25/2008 EclipseInTheMobileNetwork_ETSIv2.doc Page 13 of 17 Copyright © 2008 Harris Stratex Networks. only low priority ‘best effort’ data is discarded. the modulation rate.5 percent of the time. all rights reserved. but it will be at the expense of a lowered system gain. and hence capacity. 64 QAM. with a guaranteed bandwidth provided all of the time. 128 QAM. but is only needed to protect the link against worst-case fades throughput is that may occur for just a few minutes in a year. • When used in conjunction with QoS traffic prioritization. It interacts with other plug-ins to provide an end-to-end solution for Ethernet only.QPSK. For the rest of the year the margin is typically available not used. Some features are subject to availability. the highest modulation is typically available for better than 99. a two- fold improvement in data throughput is provided for a change from QPSK to 16 QAM. Changing to a more efficient modulation scheme will provide more link capacity.selected by an adaptive modulation engine that can handle up to 100 dB/s fading fluctuations. is used to support critical traffic with a 99. so changing to a wider channel bandwidth to achieve a higher link capacity is not an option. as the most robust modulation. 14 or 28 MHz. and hence link availability. Highest capacity. TDM only. or a combination of Ethernet and TDM payloads. Instead of using a fixed modulation rate to provide a guaranteed capacity and service availability under all path conditions. . 16 QAM. A link using robust QPSK modulation can have as much as 256 QAM 30 dB of fade margin.

and a four-fold improvement to 256 QAM. Co-Channel Dual Polarized Links (CCDP) In situations where increasing the channel bandwidth and/or increasing the modulation rate. White Paper fold improvement to 64 QAM. all rights reserved. expands to 45 Mbit/s Eclipse CCDP using 256 QAM adaptive modulation. This typically means that on the original 10 Mbit/s link. the other the horizontal. • E1 connections are dropped in user-specified order when link capacity is reduced. continues to get through when path conditions are poor. such as voice and video. It doubles wireless capacity Investments in over the same channel. The adaptive modulation engine ensures that the highest throughput is switching is always provided based on link quality. One modulation. will not provide the capacity needed. this can be increased by a existing channel factor of 4:1 to provide an overall improvement of 8:1. Note that while adaptive modulation can also be used on PDH links and combined PDH and Ethernet links. and the service provisioning provided by any MPLS or PBB-TE network overlay. existing traffic is unaffected during a change to a discarded. During a change to a lower modulation. Outside these conditions ‘best effort’ lower priority traffic.doc Page 14 of 17 Copyright © 2008 Harris Stratex Networks. Note that in many instances the link parameters that supported the original system gain can be retained. For maximized using example. higher modulation. 45 Mbit/s throughput is provided for 99. the antenna sizes and Tx power used for an original QPSK link on a 7 MHz channel (10 Mbit/s / 5xE1) are unchanged when operated on 256 QAM using Modulation adaptive modulation. enjoy data bandwidths that can be up to four times the guaranteed bandwidth. a 7 MHz RF channel supports just 10 Mbit/s using QPSK. An adaptive coding capability will provide options to trade off throughput against system gain. CCDP provides an answer.5% of the time. hitless for traffic that is not Modulation switching is hitless. 720 Mbit/s Link Aggregated Mixed Mode CCDP Terminal 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. . plans can be This has special significance on capacity extensions needed to support HSPA 3G and 4G base stations where figures in excess of 50 Mbit/s are forecasted for mid-city sites. then doubles to 90 Mbit/s using CCDP. When coupled with adaptive modulation. • Ethernet connections enjoy real synergy through the QoS awareness on the GigE plug-in. For example. All high priority traffic. and Cross Polarized Interference Cancellation (XPIC) is used in the radios to ensure any interference between the channels is eliminated. and restored when capacity is restored. Similarly. remaining higher priority traffic is not affected. and adaptive The CCDP option provides two parallel communication links on the same RF channel. such as email and file transfers. Some features are subject to availability. link uses the vertical polarization. Figure 12. unlike Ethernet there is no QoS synergy on PDH connections. A single twin-feed dual polarized antenna is installed at each end of the link.

. • The capacity on each link can be used for Ethernet. Some features are subject to availability. 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. • Modulation options extend from QPSK to 256 QAM. TDM. For mobile backhaul networks. • Capacity maximums (both links) extend to 720 Mbit/s. 200xE1. • If both links are configured for Ethernet. and the fitting of plug-in XPIC radio access cards. but must retain support for legacy PDH interfaces at 2G and 3G base stations. the two traffic streams can be L1or L2 link- aggregated onto a single user interface. The HSX pseudowire solutions comply with industry-standard IETF and MEF 8 PWE3 options for: • TDM over IP/MPLS (CESoPSN – structured E1) • TDM over IP/MPLS (SAToP – unstructured E1) • TDM over Ethernet (CESoETH – MEF 8) • ATM pseudo wires (legacy R99 traffic) • HDLC and PPP pseudo wires (CDMA market) An industry-standard solution is essential to ensure network-wide connectivity and interoperability.White Paper Figure 13. 720 Mbit/s Link Aggregated RSTP CCDP Ring Node Adding CCDP operation to an existing Eclipse link simply requires installation of the parallel link. replacement of the existing antenna with a dual-pol antenna. all rights reserved. two CCDP link pairs (four links in total) can support up to 1. • CCDP links can be 1+1 protected (hot-standby or diversity).doc Page 15 of 17 Copyright © 2008 Harris Stratex Networks.4 Gbit/s. or 4xSTM1. Where even higher capacity links are needed. pseudowires will be needed where a PDH network is replaced by an Ethernet or IP/MPLS based network. fixed or adaptive. For example. or operated in a Super- PDH ring. Pseudowires Psuedowires are used to transport a native service over a packet switched network. or a mix of TDM and Ethernet (Liquid Bandwidth). • Eclipse CCDP configurations cover channel bandwidths from 7 to 56 MHz. three or four links can be installed.

• For pseudowire-connected legacy base stations an external IEEE 1588v2 slave/translator is used to generate the clocking source. Ethernet. for converged services the options most preferred are IEEE 1588v2 Precision Time Protocol (PTP). The resultant dynamic bandwidth allocation between the different aggregated services means maximum use is made of available capacity on the trunk. Standards-compliant MPLS and PBB-TE solutions are available from Harris Stratex to ensure interoperability with those from leading core network vendors. an alternative end compatibility clocking source is required. IP/MPLS and PBB-TE The traffic engineering capabilities of IP/MPLS are well established. Others options must be have indicated that they will likely incorporate both. which when aggregated with HSPA data can provide 4:1 bandwidth efficiency gains. when there are no voice calls. so is well positioned to support the evolution of mobile networks from mixed mode to all IP transport. all rights reserved. Bandwidth Optimization and Traffic Aggregation Optimization and aggregation and can provide dramatic capacity efficiencies on backhaul connections where. . Within a PDH backhaul network.8262) is fully promulgated. and as these networks evolve it is being pushed further towards the network edges. with synchronous in the core.5 Node Bs that incorporate an IEEE 1588v2 slave.doc Page 16 of 17 Copyright © 2008 Harris Stratex Networks. the full trunk bandwidth is available for data traffic. 1588 in the Metro and access portions of the network. • While optimization and aggregation can apply on existing TDM network connections. As the number of voice calls increases. Some features are subject to availability. • Eclipse currently supports transparent transmission of IEEE 1588v2 frames . including the adoption of pseudo wires. 2G and 3G circuit-switched connections are converted to packet-based data and aggregated (multiplexed) using the aggregation capabilities of a layer 2 Ethernet switch. 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. • Optimization and IP/Ethernet conversion is available for Abis and lub (3G R99). While it is apparent that operators will choose implementations on converged based on their particular legacy infrastructure and future plans.Eclipse can be direct-connected to Ethernet-ready Rev. E1s are used as the BTS clocking source to provide base station synchronization. with service priority issues handled by the QoS engine. and consequently offers less complex and lower cost traffic engineering solutions compared to MPLS. PBB-TE. for example. It is the prominent core network protocol for converged networks. • A synchronous GigE option will be provided when the standard (G. they enable particularly cost effective gains on Ethernet or IP/MPLS networks where a common transmission pipe is shared for aggregated services. the bandwidth for data is dynamically backed off. and IEEE standards. is optimized for Metro Ethernet networks. It provides predicable and robust routing and QoS differentiators for multiple service levels and customer groups. White Paper Network Synchronization To ensure end-to. and synchronous networks. unlike MPLS. Traffic aggregation is supported on the Eclipse Ethernet interfaces. If replaced by an all-Ethernet backhaul network. synchronization Some operators have indicated intent to use IEEE 1588v2 or synchronous Ethernet. Harris Stratex solutions will support both options. compliant.

84 Mbit/s (STM0) and 155. VPLS Virtual private LAN service. QoS Quality of service. Transmission rates range from 51. Denotes a multipoint or point-to-point Ethernet connection over a host network. and in a carrier Ethernet Network is a physical 10. or 10 Gbit/s Ethernet interface. Evolving standard for 4G mobile networks.52 Mbit/s (STM1) through to 10+ Gbit/s. Function is enabled on an eLSR. SLA Service level agreement. VLAN Virtual LAN. PDH Plesiosynchronous digital hierarchy. 7/25/2008 EclipseInTheMobileNetwork_ETSIv2. Function is enabled on an LSR. 100. NGN Next generation network. and transparent LAN services. It is always provided by the service provider. HSPA High speed packet access. RSTP Rapid spanning tree protocol. LSR Label switch router. and Ethernet Internet Access. layer 1 level/interface. Applicable to private lines. TM RWPR Resilient Wireless Packet Ring. HSDPA High speed downlink packet access. virtual private Interface to the customer network in an MPLS network. VPN Virtual private network. or 1000 Mbit/s.wi-fi. Some features are subject to availability. EVC Ethernet virtual connection.White Paper Glossary CCDP Co-channel dual polarized. MEF Metro Ethernet Forum. An enhanced modulation technique designed to provide data rates up to 384 Kbps.1Q tagging mechanism. Asynchronous multiplexing scheme in which multiple digital synchronous circuits run at slightly different clock rates. HSUPA High speed uplink packet access. WiFi is a trademark of The Wi-Fi Alliance (www. EMS Element management system. all rights reserved. Provides the core function of MPLS label switching and functions as an MPLS Provider (P) node in an MPLS network. Phy Physical. SDH Synchronous digital hierarchy. XPIC Cross-polarized interference cancellation. EDGE Enhanced data rates for GSM evolution. PE node Provider edge node in an MPLS network. TDM Time division multiplexing. E-Line Denotes an Ethernet point-to-point service.doc Page 17 of 17 Copyright © 2008 Harris Stratex Networks. UNI Universal network interface. E-LAN Denotes an Ethernet multipoint service. such as SDH. Interoperability brand behind the IEEE 802. Provides the edge function of MPLS label switching and functions as an MPLS Provider Edge (PE) node in an MPLS network. eLSR Edge label switch router. CE Customer edge. IEEE 802. SFP Small-form-factor pluggable. P node Provider node in an MPLS network.16 Metropolitan Area Network standards. . WiMAX Worldwide Interoperability for Microwave Access. Applicable to multipoint L2 VPNs. Multiple low-speed signals are multiplexed to/from a high-speed channel. The physical port between the customer and service provider. with each signal assigned a fixed time slot in a fixed rotation. LTE Long term evolution. WiFi Wireless Fidelity.