Where copper gets carrier grade

T E C H N O L O G Y W H I T E P A P E R Where copper gets carrier grade High-speed Carrier Ethernet and E1/T1 on the same copper line In the competitive market place, end users are increasingly demanding fixed and mobile business services. As a result, operators are looking for ways to migrate to simpler, more cost-effective Carrier Ethernet access networks. The aim is to reduce capital and operational expenditures, as well as increase bandwidth and service options to customers. Fiber-connected business and cell sites have the inherent capacity and flexibility to meet this demand already. However, the remaining challenge is to make Carrier Ethernet services equally available to the vast majority of business and cell sites that do not have direct access to fiber. Another challenge is that many business and cell sites are currently connected by copper E1/T1 backhaul facilities, which do not support today s bandwidth requirements. Fast growing data communication traffic has resulted in substantially higher peak throughput, as well as high backhaul bandwidth demand. The answer to these challenges is a technology that simultaneously provides higher capacity broadband Ethernet services and TDM on the existing copper infrastructure. With Ethernet in the First Mile (EFM) over copper, operators can offer high bandwidth to business customers and backhaul facilities for cell sites. By adding a unique technology that transports TDM simultaneously with Ethernet over the same copper line, the operator can also facilitate a smooth migration from a TDM to an all-packet network.

Table of contents 1 1 Ethernet First Mile over Copper 1 1.1 Business access challenges 2 1.2 What is Ethernet in the First Mile? 3 1.3 EFM over copper - the physical layers 4 1.4 EFM over copper - multi-pair aggregation 5 1.5 2BASE-TL versus multiple E1/T1 circuits 6 1.6 Key requirements for EFM solutions 8 2 Hybrid Packet/TDM solution 8 2.1 Backhaul challenges 10 2.2 What is the Hybrid Packet/TDM solution 11 2.3 Why introduce the Hybrid Packet/TDM solution 12 2.4 Cost savings 12 3 Alcatel-Lucent EFM solutions for copper networks 12 3.1 Alcatel-Lucent 1531 Copper Line Access Switch (1531 CLAS) EFM Platform 13 3.2 Alcatel-Lucent 1532 Copper Line Access Switch (1532 CLAS) Hybrid Packet/ TDM Platform 14 3.3 Conclusion 15 4 Authors

1 Ethernet First Mile over Copper 1.1 Business access challenges As a dominant force in LANs for decades, Ethernet has invaded the service provider network as the preferred Layer-2 transport technology for core and metro networks. Ideal for next-generation IP and VPN services, Ethernet yields simplicity for easy deployment, flexibility for a variety of services, and lower cost for large-scale networks. In the access network today, Ethernet is most often transported over optical fiber while large parts of the copper access network are still limited to traditional TDMbased technology. To be sure, Ethernet services over copper are quickly becoming a significant contributor to the telecom service provider s bottom line. They not only enable the cost-effective use of their copper infrastructure for new customers but also enhance the carrier s service offerings. Currently, over 80 percent of customer sites are still connected via copper; only a few large companies reap the benefits of symmetrical high-speed services. According to Vertical System Group, less than 20 percent of businesses with more than 20 employees had access to fiber in 2007. Figure 1. Fiber availability to commercial buildings with 20+ employees U.S. Europe Fiber 15.3% Fiber 10.7% No fiber 84.7% No fiber 89.3% Source: Vertical Systems Group ENS Research Program Although fiber continues to be deployed more and more, it will still take many years, as well as a much larger investment, before most business customers are offered direct optical connectivity. In the meantime, however, with the possibility of offering the same service on copper, the high-speed packet services can be extended to many small- and medium-size businesses. Before deciding whether to re-use the existing copper infrastructure, or to invest in microwave or fiber, service providers must take deployment costs, implementation time and the need for capacity into consideration. In order to limit the immediate capital and operational expenditures inherent in fiber deployment, many service providers make the choice for a copper-based solution, better leveraging their existing assets. Until now, there have been few cost-effective solutions other than traditional E1/T1 and these are only cost-effective when the incumbent operator is regulated to make them cost effective. The aggregation of E1/T1 solutions has been frequently used to increase capacity in a linear way, but the bandwidth increase is limited and the cost of this solution is high. However, with the ratification of the IEEE 802.3ah Ethernet in the First Mile (EFM) standards, the economics of access have Where copper gets carrier grade Technical White Paper 1

changed. The standards offer a new technology for delivering high-bandwidth Ethernet services over copper facilities. With this technology, service providers can accelerate business access deployment to copper-connected sites with speeds typical of fiber services. Now more than ever, it is critical to consider the economic ramifications of every deployment decision. The need to win more customers and get more revenue from existing customers without large investments is critical to the success of every provider. As the competition for broadband business customers heats up, the decision made about the last-mile service infrastructure could easily make or break the carrier. Access is everything. 1.2 What is Ethernet in the First Mile? 1.2.1 The Ethernet in the First Mile Project (IEEE 802.3ah) The demand for Ethernet in the carrier network was realized in the IEEE 802.3 Ethernet standards body, which created the 802.3ah Ethernet in the First Mile (EFM) task force. The task force enhanced existing Ethernet standards targeted at carrier concerns about the deployment of Ethernet in the access network. To do this, it quickly focused on the following four areas all of which required new standards development: 1. New optical Ethernet physical layers with the reach, bandwidth, and environmental characteristics for outside plant deployments. 2. An Ethernet passive optical network (E-PON) technology for residential deployment of triple play (voice, video and data) services. 3. Operations, administration, and maintenance (OAM) functions to simplify monitoring and troubleshooting geographically dispersed Ethernet networks. 4. New physical layers for Ethernet over standard copper cables for universal broadband Ethernet coverage. 1.2.2 EFM over copper cables In the fourth area, the task force developed two new Ethernet physical layers in order to address two very different and important market segments. The first technology, 2BASE-TL, is a long-reach Ethernet-over-copper technology, which focuses on high-bandwidth symmetric services for business customers from Central Offices (COs) or remote terminals. 2BASE-TL is the natural upgrade and replacement for today s E1/T1. The EFM long-reach solution is based on G.SHDSL. The second technology, 10PASS-TS, is a short-reach, high-bandwidth asymmetric technology targeted for in-building or FTTC deployments. The 10PASS-TS technology permits more bandwidth options for residential or business access. Both 2BASE-TL and 10PASS-TS enable native Ethernet frames to move across existing voice-grade copper pairs in carrier access networks. The EFM shortreach solution is based on VDSL. The EFM technologies cover the full spectra of copper access deployment possibilities from shortreach to long, from business to residential. With regard to both rate and reach, 2BASE-TL and 10PASS-TS combine to blanket the chart as illustrated in Figure 2. 2 Where copper gets carrier grade Technical White Paper

Figure 2. EFM family rate/reach applicability Speed Multipair 2BASE-TL 10PASS-TS 2BASE-TL Distance With 10PASS-TS, EFM can offer very high rates on very short loops, reaching as high as 100 Mb/s in asymmetric mode, and 50 Mb/s in a more symmetric mode. The technology is targeted at 1,500 m (5,000 ft) and below for Customer Serving Area (CSA) distances and is aimed at true triple-play residential services, with asymmetric, very high-bandwidth service potential. For its part, 2BASE-TL is focused on delivering symmetric services to business customers from COs and remote terminals. It supports service delivery out to CSA distances - 2,700-3,600 m (9 to 12,000 ft) and beyond. With a maximum symmetric rate of 5.7 Mb/s per pair, 2BASE-TL delivers highbandwidth Ethernet services over just a single pair. Businesses can now be reached with symmetric services of 10 Mb/s and higher, on as few as two pairs of copper, as 2BASE-TL continues to be rolled out as the next-generation replacement for traditional E1/T1 services. 1.3 EFM over copper - the physical layers The EFM standard leverages the latest DSL layers, as defined by the International Telecommunications Union (ITU), as the physical layers for Mid-Band Ethernet. By using these standards, IEEE 802.3ah continues to benefit from the high volume of DSL chipsets, while significantly improving upon the original silicon by defining new and efficient mechanisms for Ethernet transport. 1.3.1 2BASE-TL 2BASE-TL is based on the same physical layer as G.SHDSL.bis (also known as G.991.2.bis or E-SHDSL). G.SHDSL.bis can run up to 5.7 Mb/s on a single pair. With such high-speed symmetric access, customers can be offered a true 10 Mb/s Ethernet service on as little as 2-pair. New chipsets are also making it possible to reach up to 15 Mb/s per copper pair. 2BASE-TL and G.SHDSL.bis increase the bandwidth potential of SHDSL in two dimensions. First, a second constellation (or symbol encoding) is allowed, which increases throughput by 33 percent without affecting the spectral properties of SHDSL. This additional, higher constellation cannot be used on the longest loops, but does provide a spectrally free increase in throughput on loops up to 3 km (10,000 ft), depending on the noise environment. Second, 2BASE-TL and G.SHDSL.bis increase the frequency (number of symbols per second) as compared to SHDSL, thus allowing even more throughput. 1.3.2 Spectral compatibility and international application One of the key issues with any technology operating over copper cables is the effect of the technology on other services. Copper cables are deployed in binder groups of tens of pairs, which are bound together in a common outer sheathing. When a service is deployed on a pair, it creates noise that affects other pairs in the binder group and sometimes in nearby binder groups as well. The impact of a technology on other services in a binder group is measured by its spectral compatibility. Where copper gets carrier grade Technical White Paper 3

No hard and fast rule exists as to what counts as too much noise. In different parts of the world, different kinds of cables are deployed; the loops are of different lengths and the deployment practices are different. Each country is therefore free to define its own spectral compatibility guidelines for services deployed in its telecommunications infrastructure. EFM technologies (2BASE-TL and 10PASS-TS) are deployable throughout the world. They can operate under different spectral guidelines depending on where they are deployed. 1.3.3 Carrying Ethernet packets over copper A long-standing tradition in Ethernet is that the method for carrying the actual frames over the cable must have low overhead and be utterly resilient to false packet acceptance (FPA). FPA refers to the probability that, when a corrupted frame is received, the corruption is not detected. Mid-Band Ethernet technologies use a novel encoding scheme called 64/65-octet encoding where there is one overhead byte for every 64 bytes of data. The scheme is highly efficient, which is vital for access technologies that must adapt to the environment in order to deliver the highest possible speed, given the existing outside plant conditions. Additionally, 64/65-octet encapsulation includes measures to improve FPA acceptance results of traditional DSL encoding. DSL physical layers generally operate in modes that yield a bit-error rate of 10-7. Traditionally, Ethernet technologies (and the IP layers above them) have been built upon an architecture where FPA cannot, in probability, occur. To achieve FPA performance acceptable for Ethernet and IP delivery, the 64/65-octet layer appends every frame (or fragment) with a Cyclic Redundancy Check (CRC) in addition to the Ethernet Frame Check Sequence (FCS). The combination of these two error-checking codes practically eliminates the possibility of FPA, thus maintaining the historical reliability of Ethernet. These benefits result in a more efficient and reliable access network technology. 1.4 EFM over copper - multi-pair aggregation EFM copper technologies introduce a novel mechanism for using multiple copper pairs in order to deliver additional bandwidth to the customer. Developed in 802.3ah, the loop aggregation techniques are able to optimize the utilization of the set of lines as well as add and remove pairs to and from the aggregate. 1.4.1 How it works IEEE 802.3ah employs a loop aggregation technique for Ethernet optimized for copper access. Figure 3. Loop aggregation technique Media access control (MAC) Reconciliation Traditional Ethernet layer MII Rate matching Loop aggregation New Ethernet layer 64/65-octet encapsulation Physical layer 64/65-octet encapsulation Physical layer Existing ITU physical layers The loop aggregation techniques of IEEE 802.3ah are both simple and powerful. Frames are passed to the loop aggregation layer from the higher layer, where they are fragmented and distributed across the loops within the aggregate. When transmitted across the individual loops, a fragmentation header is prepended (see Figure 4), which includes a sequence number and frame markers. This header is used by the receiver to resequence the fragments and to re-assemble them into complete 4 Where copper gets carrier grade Technical White Paper

frames. To allow vendor differentiation, the algorithm for partitioning the frames over the loops is not specified. However, it must obey certain rules in that fragments must obey size constraints, and the loops in an aggregate must obey rate and differential delay constraints. As long as the loop aggregation algorithms conform to these constraints and restrictions, any fragmentation algorithm can be handled by the re-assembly process, yielding a very flexible and interoperable solution. Figure 4. Fragmentation header is prepended Frame Loop aggregation fragmentation FH Frag-1 FH Frag-1 FH Frag-1 Loop aggregation fragmentation Frame FH = Fragment header SOP EOP SeqNum SOP = Start of packet flag EOP = End of packet flag SeqNum = Sequence number 1.4.2 Automatic resiliency for the most demanding customers and applications In addition to the efficiency and performance benefits of 802.3ah, loop aggregation has the added benefit that it s automatic. Pairs can come and go, and the Ethernet interface remains operational only the available bandwidth is affected. New pairs can be wired up and automatically joined to the aggregate group with no additional configuration, realizing the plug-and-play potential of Ethernet. The resiliency of 802.3ah loop aggregation can satisfy the most demanding business customers and support any application. When a pair fails, that pair is detected and removed from the aggregate in just a few milliseconds. Established Voice-over-IP calls remain operational, and the callers don t even notice that a problem has occurred. Video streams continue to play as if nothing has changed. Applications and their users are unable to detect that one of the pairs has failed, except by the loss of some bandwidth. And when that pair comes back online, it is seamlessly added to the aggregate. That again goes unnoticed by the applications and users. This makes IEEE 802.3ah the most suitable technology for the business services of today and tomorrow, where unreliable, best-effort delivery is simply not enough. 1.5 2BASE-TL versus multiple E1/T1 circuits Carriers have multiple options for reaching the business customer with business quality symmetric services. For example, business customers have traditionally been served with E1/T1 circuits. The natural integration of E1/T1 circuits into a SDH/SONET network, as well as the relative ubiquity of the service offering, has traditionally made E1/T1 circuits one of the most popular choices for access technology. As the bandwidth needs of customers grow, it is natural to consider multiple E1/T1s as a higher speed access solution. This section examines the differences between multiple E1/T1s and 2BASE-TL. Where copper gets carrier grade Technical White Paper 5

1.5.1 Simplicity and consistency Perhaps the most significant benefit of EFM is the simplicity it brings, and the resultant savings in capital and operating expenses. As true Ethernet technologies, 2BASE-TL and 10PASS-TS can yield immediate and large operational savings. Additionally, 2BASE-TL and 10PASS-TS yield consistent next-generation architectures across any medium. The only difference between customers connected via fiber and customers connected via twisted pair is the potential bandwidth available to the customers. The services and management remain consistent between headquarters and branch office locations of the customer, allowing the provider to capture the whole customer with a differentiated, high-margin service set. 1.5.2 Access architecture and costs Another significant difference between a multi-e1/t1 solution and 2BASE-TL is the access architecture and solution costs. 2BASE-TL requires plain copper access lines for the customer service whereas multi-e1/t1 solutions require E1/T1 interface; one requires a facility (copper), the other requires a service (E1/T1). The exact cost of these different access mechanisms is highly dependent on geography. However, a dry copper pair is almost universally much less expensive than an E1/T1. When using multiple lines for higher speed services, the difference in cost is even more dramatic. To offer a 10 Mb/s service using multi-e1/t1 lines would require 7 circuits. To offer a 10 Mb/s service using 2BASE-TL would require 2 to 8 pair (depending on the distance and noise environments). It s clear that the carrier offering a 10 Mb/s service using 2BASE-TL could do so at significantly lower cost than the carrier offering that service using multiple E1/T1 lines. Additionally, even when E1/T1s are required for physical or service connectivity, the same economic benefits as 2BASE-TL can be realized while delivering a native Ethernet or E1/T1 service. By applying the latest technological advances, E1/T1 services can be delivered over leased copper just as Ethernet services can be delivered; and just as more Ethernet bandwidth can be delivered more economically, more E1/T1 bandwidth can also be delivered more economically. 1.5.3 More bandwidth and new services over the same copper pairs A multi-e1/t1 approach provides the carrier with a mechanism to deliver a higher bandwidth solution at a significantly higher cost and complexity than 2BASE-TL. The carrier can continue its traditional E1/T1 service offerings with simply a larger pipe to the customer. On the other hand, 2BASE-TL allows the definition of a new and differentiated service offering with significantly higher bandwidth and revenue potential. Instead of a 6 Mb/s service over three E1 or four T1 lines, 2BASE- TL can provide more than 22 Mb/s (with the new chipset up to 60 Mb/s) over four access lines. The potential bandwidth differences allow a new and exciting range of services that would be difficult or impossible over legacy technologies, such as E1/T1. In addition, 2BASE-TL equipment delivers plug-and-play provisioning, and is less costly in capital and operational costs than multi-link E1/T1 alternatives. 1.6 Key requirements for EFM solutions Adhering to the IEEE 802.3ah 2BASE-TL standard is just the first requirement for providing a carrier-class EFM solution. Additional requirements must be met to reach the full potential of 2BASE-TL products. For example, full Layer-2 switching with quality of service (QoS) is necessary for carriers to truly differentiate their service offerings, and carrier-grade reliability is expected in order to achieve the same reliability and dependability as today s TDM-based services. 6 Where copper gets carrier grade Technical White Paper

1.6.1 Bandwidth in copper access networks is precious and requires intelligent protection The ability to deliver high-margin, differentiated services is fundamental to a carrier s ability to generate meaningful new revenue, to keep customers happy, and to compete for new business opportunities based on performance and quality rather than price. While some believe that an overabundance of bandwidth can serve as a means to achieve QoS, they are missing a key opportunity to leverage advances in Ethernet platforms capable of delivering carrier-grade Service Level Agreements (SLAs) based on a flexible service framework. And bandwidth, though sometimes abundant and inexpensive in the core, is still quite precious in the access network. QoS technology provides a method for categorizing traffic and for ensuring that particular categories of traffic will always flow across the network at their specified bandwidth, latency, and jitter service levels, regardless of competing demands. It can also provide guarantees of bandwidth, error rate, and many other characteristics. For example, let s take a fully capable Layer-2 Ethernet switch that supports two independent QoS parameters for every packet that enters the system. The first parameter is the Class of Service (CoS), which is a priority level that controls the latency of the packet through the switch. The second parameter is the discard eligibility that controls the probability of the packet being discarded under congestion. An ingress classification and policing process determines the priority and discard eligibility of each frame. A mapping of packet CoS markings (such as 802.1p-bits or IP DSCP bits) to CoS values can be configured for each port. Then, when provisioning services over that port, the CoS values and service identifiers are mapped into traffic flows. These traffic flows are given bandwidth parameters, and the discard eligibility for each packet is based upon those parameters and the amount of traffic entering the system in that flow. Each customer can then be treated and served independently. The flow classification stage permits multiple traffic flows for each customer within each service. This allows, for example, different bandwidth controls for a VoIP application compared to an e-mail application from the same customer. Packets are first classified into services to determine the forwarding rules for the packets, and then they are mapped into flows to control the priority and bandwidth of any data stream. Per flow discard eligibility parameters are used to treat traffic fairly during periods of congestion. The result is a solution that overcomes the limitations of ATM-based DSL access multiplexers (DSLAMs) and traditional Ethernet switches with a QoS solution that integrates and interoperates for end-to-end service quality. 1.6.2 The benefits of QoS in Mid-Band Ethernet networks SLAs, when provisioned on a per-port and per-service basis, can be used to tier services based on a flexible QoS and CoS framework. All services can then be converged over a single physical connection with their own SLA. Another advantage of using a QoS and CoS framework is the ability to help carriers deliver stickier services. Higher levels of customer satisfaction lead to stickier services, meaning business customers are less likely to seek alternative carriers for their data service requirements. Keeping customers happy and loyal is a key challenge for carriers today and in the future as they face renewed competitive threats from alternative carriers for higher speed data access, and from cable providers for lower speed access. Ethernet access can play a key role in maintaining customer loyalty by enabling carriers to deploy a wider range of services, delivering them more quickly and efficiently, and tailoring them to more closely match specific customer needs. Where copper gets carrier grade Technical White Paper 7

Today s Ethernet QoS and CoS capabilities offer substantial benefits to carriers and their customers alike. Improved application performance, along with support for new applications and services, such as point-to-point private line, point-to-multipoint, multipoint-to-multipoint, transparent LAN, and Internet access, are all possible when QoS and CoS are deployed network-wide, beginning with the edge device. 1.6.3 Ensuring complete business-grade resiliency Business access services require a level of resiliency beyond that of residential services. Resiliency, in order to be effective, must exist at every layer in the network hierarchy, as the service is only as resilient as its weakest link. Carriers today expect and demand new product offerings to provide carrier-grade availability of five 9s. Today s IP/Ethernet products now carry time-sensitive traffic such as voice and video, in addition to critical business applications. Achieving the same reliability and dependability as today s TDM-based services is a top requirement of carriers. EFM products should be designed with true business-grade resiliency rather than a semi-redundant system. Resiliency is required across every aspect of the system and network. The following are necessary to provide the system and network redundancy that carriers need to satisfy the most demanding service up-time requirements: bonded copper ports to their customers; stackable nodes for system resiliency; spanning tree protocol and link aggregation for network resiliency. 1.6.4 Key features of a business-grade resilient solution Adhering to the IEEE 802.3ah 2BASE-TL standard is just the first requirement for providing a carrier-class Mid-Band Ethernet solution. Additional requirements must be met to reach the full potential of 2BASE-TL solutions, including full Layer-2 switching with QoS, carrier-grade reliability, a scalable architecture, and a low initial cost to achieve a fast return on investment (ROI). The key benefits of a business-grade solution are: Full Layer-2 Ethernet switch with spanning tree and rapid spanning tree protocols In multi-stack solutions: a robust architecture preventing any single points of system failure using redundant interconnections with distributed management control agents distributed bonding of copper pairs across multiple units for ultimate business-class resiliency, simplified pair management, and no stranded ports Complete carrier-grade OAM in the Ethernet network Controls to limit unicast and multicast bandwidth Controls per-customer MAC address resources. With effective QoS and CoS capabilities, carriers can leverage Carrier Ethernet as a foundation for a new class of data services by layering multiple SLA-based services over each physical connection. These services, based on the performance and bandwidth requirements demanded by high-value business customers, will be fundamental to a carrier s ability to cost-effectively compete for last-mile business customers on performance, not just price. 2 Hybrid Packet/TDM solution 2.1 Backhaul challenges E1/T1 has been the gold standard for the transport of both voice and synchronization, but E1/T1s lack the bandwidth and reliability required for the burgeoning broadband data demands. Mobile operators migrating to 3G or 4G cellular networks are offloading their broadband data onto high- 8 Where copper gets carrier grade Technical White Paper

bandwidth Ethernet backhaul. However, they are also leaving E1/T1s in place for voice services and synchronization, while they search for a way to get all the bandwidth and cost advantages of Ethernet without jeopardizing the benefits and performance of E1/T1s for voice and physical layer timing. Figure 5 below illustrates the growth in data traffic and the reduced revenue per bit. Figure 5. Increased data traffic and reduced revenue per bit Traffic Voice dominant Cellular operator revenue and traffic de-coupled Revenues Data dominant Time In today s data dominant market, E1/T1 leased lines cannot solve this challenge they lack the economics and robustness to meet the mobile operator demands. However, there are many elegant solutions for the carrier to accomplish this goal when high-speed fiber optics facilities are available. Unfortunately a considerable part of the cell sites are served only by copper E1/T1 facilities. Even when upgraded to Ethernet over bonded copper (with considerably more bandwidth than E1/T1s), speeds across these last-mile copper facilities are much lower than fiber in the 10 Mb/s to 50 Mb/s range. That means latency, jitter and timing become issues in Ethernet-based setups over last-mile copper facilities. The Hybrid Packet/TDM solution offers a high quality, highly resilient, immediately synchronized timing mechanism. In addition, it provides fundamental elimination of jitter and bit error multiplication, thus ensuring a synchronization signal that meets E1 and T1 standards. Achieving efficient utilization of the backhaul network requires consolidation of TDM and broadband data traffic onto a single backhaul facility. The Hybrid Packet/TDM solution does that by handling TDM traffic natively retaining all of the requisite native TDM attributes and physical layer timing delivered by E1/T1s. With E1/T1s, loss of a single pair will result in loss of service and cell tower synchronization. By contrast, the Hybrid Packet/TDM solution uses a multi-pair transport facility that carries both native TDM traffic as well as native Ethernet, and spreading and prioritizing voice timeslots across the bonded copper pairs, while dynamically allocating the remaining bandwidth for the broadband Ethernet traffic. With the Hybrid Packet/TDM solution, the service provider reaps all the benefits of TDM for synchronization and voice, while achieving the bandwidth and economic returns of Ethernet. Circuit emulation, which can be enabled over any packet access network including a copper-based one, unfortunately suffers from poor TD performance, long synchronization times, and substantially worse bit error performance. The Hybrid Packet/TDM solution also does not have the problem of stranding bandwidth through static allocation of traffic streams over specific copper pairs as do dual-bearer mode technologies. Where copper gets carrier grade Technical White Paper 9

2.2 What is the Hybrid Packet/TDM solution The Hybrid Packet/TDM solution offers flexible, resilient transport of both Ethernet and TDM traffic by delivering native TDM traffic simultaneously with native Ethernet traffic over the same bonded copper connection. The technology lets operators transition their mobile backhaul networks from Time Division Multiplexing (TDM) to Ethernet, while providing their customers the bandwidth they need and line quality they expect. The technology delivers any mix of native TDM and native Ethernet over bonded copper and is an alternative to PWE3 (Pseudowire Emulation Edge to Edge) transport of voice services over the copper access networks, while being interoperable with PWE3 over an optical packet core. The solution overcomes the limitations of pseudowire on copper access networks by: partitioning a bonded copper connection into bits that carry TDM traffic and bits that carry Ethernet traffic, and transporting TDM natively rather than in a packet format. Figure 6. Hybrid Packet/TDM solution for mobile backhaul Cell site Native TDM + Native Ethernet + Bonded copper = Native Ethernet + native TDM over bonded copper The Hybrid Packet/TDM solution is not intended as a replacement for pseudowire. It complements and interoperates with a pseudowire solution by extending native TDM transport further into the network, in particular over the access network where bandwidth, congestion, and noise can significantly disrupt pseudowire performance. The solution can also complement pseudowire over high-capacity fiber optic transport. The solution offers a flexible bandwidth multiplier for resilient transport of both Ethernet and TDM traffic over multiple pairs of copper, and delivers the significantly higher bandwidth required for a carrier s typical suite of mobile, DSLAM backhaul, Wi-Fi and WiMAX backhaul applications. Carriers will be able to prioritize and dynamically allocate bandwidth to a changing mix of voice and data traffic on their aggregated backhaul transport facilities. The technology provides a significantly bigger pipe than is available on traditional copper, with enough flexibility, capacity and resiliency to meet service demands now and in the future. The Hybrid Packet/TDM technology is optimized for 2G, 2.5G or 3G on NxE1/T1 TDM and Asynchronous Transfer Mode (ATM)/TDM backhaul networks, enabling a seamless transition to all- Ethernet RAN and 4G networks. As carriers evolve their networks from TDM-based 2G networks to all-ethernet 4G, they will require optimized performance and a seamless transition. The Hybrid Packet/TDM solution allows a mobile operator to tune its TDM-versus-Ethernet requirement at any stage of the migration process. The operator then gains complete control of the TDM-packet evolution, realizing immediate economic gains. 10 Where copper gets carrier grade Technical White Paper

With the Hybrid Packet/TDM solution, 100 percent of the bandwidth on the backhaul facility is available at all times, regardless of the traffic mix currently on the network. Operators can use proven TDM timing and synchronization while also carrying Ethernet traffic natively. The ability to handle Ethernet in native format is a key benefit, since the data backhaul capacity can be seamlessly increased by simply reassigning unused voice bandwidth to the data stream at any time. 2.3 Why introduce the Hybrid Packet/TDM solution 2.3.1 Reservation of bandwidth for TDM The Hybrid Packet/TDM solution partitions a bonded copper connection into bits that carry TDM traffic and bits that carry Ethernet traffic. The TDM bits are reserved and cannot be interfered with by any Ethernet traffic. That means the delay and jitter of TDM remain constant, regardless of what else is going on in the network. TDM never has to wait in queue behind other (potentially many and potentially large) frames. The absence of the packetization effect results in a first-rate performance with regard to latency and jitter and makes the Hybrid Packet/TDM solution superior for transporting TDM in the access copper network. This advantageous approach is perfect for mobile network backhaul as mobile networks have strict requirements on latency and jitter for each part of the network because of the cumulative effect of these phenomena and restricted timing budgets. 2.3.2 No BER (Bit Error Rate) multiplication The Packet/TDM solution does not suffer from the BER multiplication as each bit error is separate and does not result in a packet error (caused by the Cyclic Redundancy Check (CRC) checking on Ethernet packets), which is often the equivalent of thousands of bit errors. On new, high quality optical connections, the multiplicative effect of pseudowire to BER is practically unnoticeable because there are so few bit errors and the speeds are so high. However, in the copper access network, which is already the weak link in the BER chain, the effect can be very noticeable. The native TDM approach described in this document eliminates the packetization of TDM and delivers a high quality service. 2.3.3 TDM quality synchronization The synchronization and timing recovery aspects of pseudowire are relatively unspecified in the standards and left to the implementations where some proprietary form of adaptive or differential timing is generally used. Both adaptive timing and differential timing have some disadvantages in the copper network and therefore pose a common challenge for copper implementations. The quality of the pseudowire synchronization signal is affected by many things, including the availability of a high quality, reliable reference clock and the amount of jitter/latency on the pseudowire stream. The Hybrid Packet/TDM solution overcomes these limitations by using all of the physical copper lines as a timing source or sink. This approach ensures the highest quality clock signal by fundamentally eliminating latency and jitter on the TDM traffic and by having multiple high quality reference clocks between devices. With the approach described, each copper pair carries a physical layer synchronization signal transporting a Stratum 3 clock. This high quality, highly resilient, immediately synchronized timing mechanism, along with the fundamental elimination of jitter and bit error multiplication, ensures a synchronization signal that can meet E1 and T1 standards. Where copper gets carrier grade Technical White Paper 11

2.4 Cost savings The Hybrid Packet/TDM solution delivers more bandwidth than E1/T1 and let carriers address a larger market footprint at lower operating expense. The cost synergies of using a single transport facility for all backhaul traffic are realized, and there is no need to physically reconfigure the network pairs. Any mix of TDM and Ethernet traffic can be supported on the Hybrid Packet/TDM solution backhaul facility with all of the bandwidth on the backhaul facility available at all times, regardless of traffic mix. The ability of the Hybrid Packet/TDM solution to help carriers control both capital expenses as they enter into a service situation, and operational expenses as they continue to serve the growing needs of their backhaul customers, is a compelling argument. Since the Hybrid Packet/ TDM solution does not put TDM into IP or Ethernet packets, the management costs and headaches from dealing with packetization delays are eliminated. Other solutions, such as dual-bearer systems carry voice traffic on one pair and bond the rest of the pairs to handle the data traffic. In such cases, the logistical and operational considerations are many, adding an ongoing cost to the equation. The carrier must have adjacent pairs for this to be successful. If the voice pair is moved, resiliency is lost. The juggling is constant as traffic demands change. 3 Alcatel-Lucent EFM solutions for copper networks Alcatel-Lucent offers two EFM over copper platforms; one platform the Alcatel-Lucent 1531 Copper Line Access Switch - offers Ethernet only while the other the Alcatel-Lucent 1532 Copper Line Access Switch - offers the Hybrid Packet/TDM solution. The CPEs, which are together with the platforms, are the 1521 and 1522 Copper Line IP (CLIP). Both platforms support repeaters for extended reach of SHDSL. Figure 7. Alcatel-Lucent EFM platforms for copper networks T1/E1 + Ethernet Ethernet 1532 CLAS 1531 CLAS 2BASE-TL and native TDM 2BASE-TL 1522 CLIP 1522 CLIP 1522 CLIP 1521 CLIP Alcatel-Lucent EFM platforms for copper networks 3.1 Alcatel-Lucent 1531 Copper Line Access Switch EFM Platform 3.1.1 Alcatel-Lucent 1531 CLAS The Alcatel-Lucent 1531 Copper Line Access Switch (CLAS) is an innovative Carrier Ethernet in the First Mile (EFM) platform that extends the reach of native Ethernet services to businesses that do not have access to fiber. The Alcatel-Lucent 1531 CLAS delivers 1 to 45 Mb/s of symmetrical Ethernet service over 1 to 8 pairs of existing last-mile dry copper utilizing standards-based 2BASE- TL technology effectively and efficiently bridging the existing E1/T1-E3/T3 service gap and economic disparity. With the Alcatel-Lucent 1531 CLAS, high-margin Carrier Ethernet services can now be delivered over voice-grade copper at full carrier serving area distances and beyond. 12 Where copper gets carrier grade Technical White Paper

Packaged in a stackable high-density 1RU pizza box, the Alcatel-Lucent 1531 CLAS is purposebuilt for deployment throughout the access network in a CO, controlled environmental vault (CEV), or remote terminal (RT). The fully front-accessible platform is standards-based, temperaturehardened, highly resilient and scalable. The Alcatel-Lucent 1531 CLAS supports an industryleading density of 40 pairs of 2BASE-TL per rack unit. It also offers TDRplus, an integrated Time Domain Reflectometer, for qualification and maintenance of copper loops. The Alcatel-Lucent 1531 CLAS also implements the IEEE standard for Ethernet OAM with extensions for complete remote management and control to simplify deployment and management, while maintaining full interoperability with existing Ethernet switches, routers and Ethernet ADM interfaces. In cases where maximum port density is required, Alcatel-Lucent s Virtual Node (VN) technology enables service providers to deploy multiple load-sharing 1531 CLAS systems as a single managed entity. Unlike other vendor offerings, Alcatel-Lucent s VN is managed as a single node, providing a fully-redundant architecture and industry leading port density. 3.2 Alcatel-Lucent 1532 Copper Line Access Switch Hybrid Packet/TDM Platform The Alcatel-Lucent 1532 Copper Line Access Switch (CLAS) is a Hybrid Packet/TDM platform offering a unique technology for the transportation of native Ethernet and native TDM on the same copper line. It extends the Alcatel-Lucent Ethernet First Mile (EFM) over copper portfolio - initially tailored for Ethernet business access also offering an attractive solution for mobile backhaul and other applications, where a combination of Ethernet and TDM is required. The Ethernet portion of the 1532 CLAS Hybrid Packet/TDM platform includes the same functionality and services as the 1531 CLAS (see description in section above). The Alcatel-Lucent 1532 CLAS delivers up to 480 Mb/s over 32 copper pairs and enables carriers to deliver Metro Ethernet services over the existing copper infrastructure to businesses whose application requirements fall within the bandwidth gap between E1/T1 and T3/STM-1. The 1532 CLAS is packaged in a 2RU chassis with the possibility of being equipped with up to five service plug-in modules in order to enable flexible scaling. It has full-front access, is temperature-hardened and supports an industryleading density of up to 80 copper pairs. Like the Alcatel-Lucent 1531 CLAS, the Alcatel-Lucent 1532 CLAS implements the IEEE standard for Ethernet OAM with extensions for complete remote management via open management interfaces. As a result, the Alcatel-Lucent 1532 CLAS is well positioned for integration into any OSS. 3.2.1 Alcatel-Lucent 1521 CLIP and 1522 CLIP The Alcatel-Lucent 153x CLAS is fully compatible with the Alcatel-Lucent 152x Copper Line IP (CLIP) series of cost-effective customer-demarcation devices, located on the customer premises. The 152x CLIP can also be used as a point-to-point link and can smoothly migrate to a point-to multipoint solution where the 152x CLIP CO equipment can be replaced by 1531 CLAS. The 1521 CLIP CO unit can also be reused as a CPE connected to the 153x CLAS. In addition, it can also be deployed as an EFM CPE for the Alcatel-Lucent converged Intelligent Services Access Manager (ISAM) platform. Where copper gets carrier grade Technical White Paper 13

Figure 8 illustrates the mobile backhaul application using the 1532 CLAS Hybrid Packet/ TDM platform. Figure 8. Mobile backhaul application using the 1532 CLAS Hybrid Packet/TDM platform E1/T1 transport over SDH/SONET network 4xE1, 1xFE Cell site gateway FE RNC/BSC site gateway RNC, BSC 4xE1, 1xFE 1522 CLIP DSL with repeaters 1522 CLIP GE Metro Core Base stations 16xE1, 2xFE/GE 1522 CLIP 1532 CLAS Native clock transport Native clock and data 1532 CLAS Clock recovery nxe1s E1 or T1 is sync source SDH E1/T1 transport over packet network 4xE1, 1xFE Cell site gateway FE RNC/BSC site gateway RNC, BSC 4xE1, 1xFE 1522 CLIP DSL with repeaters 1522 CLIP GE Metro Core Base stations 16xE1, 2xFE/GE 1522 CLIP 1532 CLAS Native clock transport Native clock and data 1532 CLAS Clock recovery TDM PWE3 (data) TDM PWE3 (clock) Primary reference clock (PRC) 3.3 Conclusion With the growing demand for new applications and service content in fixed and mobile networks, the bandwidth required for today s services is exploding. As a result, core networks have begun the migration from traditional TDM-based networks to service-rich IP/MPLS networks, thus ensuring lower CAPEX and lower OPEX. To truly capitalize on the potential revenue and cost benefits of the next-generation network, carriers must leverage a true packet infrastructure. For the foreseeable future, fiber has a severely limited footprint, which forces carriers to use the universal medium of copper as the predominant access method. EFM technologies of 2BASE-TL and 10PASS-TS were developed to allow carriers to utilize their existing copper infrastructure as high-speed, high-margin on-ramps to their metro and core packet networks. Native Ethernet in the access network lowers cost, simplifies deployments, and yields a more flexible network for new services, such as VoIP. To meet these market demands, Alcatel-Lucent is offering an EFM platform tailored for business access, which meets all the requirements described for a business-grade platform. 14 Where copper gets carrier grade Technical White Paper

The majority of 2G/3G base station sites are also still connected via copper. The challenge is to cost-effectively migrate this backhaul network to bridge the remaining service gap in the copper network. Ideally these copper-connected base stations should be reachable only via EFM technologies. However, as E1/T1 is still needed for a high number of base stations, a solution capable of transporting E1/T1 is also required. Until now, pseudowire has been seen as the answer, but this solution still has challenges due to the packetization of E1/T1 and timing. To solve these challenges, Alcatel-Lucent has introduced a hybrid solution capable of simultaneously transporting both broadband Ethernet services and TDM on the existing copper infrastructure, while still meeting the synchronization specification for E1/T1 defined by ITU. 4 Authors Matt Squire Hatteras Networks Per Hasselstrom Alcatel-Lucent Where copper gets carrier grade Technical White Paper 15

www.alcatel-lucent.com Alcatel, Lucent, Alcatel-Lucent and the Alcatel-Lucent logo are trademarks of Alcatel-Lucent. All other trademarks are the property of their respective owners. The information presented is subject to change without notice. Alcatel-Lucent assumes no responsibility for inaccuracies contained herein. Copyright 2009 Alcatel-Lucent. All rights reserved. CAR9718081221 (01)