Tellabs 8600 Overview

OVERVIEW

Tellabs 8600 ®

Managed Edge System Overview

OVERVIEW — TELLABS ® 8600 MANAGED EDGE SYSTEM

Introduction The aim of this document is to give the reader a basic understanding of the Tellabs ® 8600 Managed Edge System and a vision as to where the platform is developing. It covers the target market, the main applications and the component building blocks that make up the system today and in a longer term. As a prerequisite, the reader should have a basic knowledge of telecommunications and mobile networks. The document is written at a relatively high level and starts with an overview of access networks from both a wireless and wireline perspective. This is followed by a more detailed introduction to the Tellabs 8600 system and how it can operate in both of these types of networks. The final section concentrates on the specific network elements and the network management system, which is an essential and integrated part of the whole solution. Tellabs operates globally and is a leading supplier of managed access transport platforms to service providers around the world. Tellabs Tellabs has a successful record of providing managed access solutions to more than 300 customers over the past 15 years. The Tellabs 8600 system is a next-generation packet-based platform that is suitable for both access and regional networks in mobile transport and converged networks. It is attractive not only to new customers building long term network solutions but also to existing Tellabs ® 8100 Managed Access System customers wishing to extend the capabilities of their current access networks. Tellabs knows how important it is to provide our customers with a seamless migration from their current solutions as they introduce new technology and provision new services. A lot of effort has therefore been made to integrate all of the network elements under one management system. A chapter of this guide is dedicated to what this means in practice and the added value that it i t provides for the customer.

2

Table of Contents Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Network evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 The Tellabs ® 860 8600 0 Managed Edge System in next-generation networks . . . . . . . . . . . . . .4 .4 The Tellabs Tellabs 860 0 system’s benefits and role in wireless transport . . . . . . . . . . . . . . . . . . .6 .6 The Tellabs Tellabs 860 0 system in GSM and UMTS net works . . . . . . . . . . . . . . . . . . . . . . . . . . .7 Customer cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 The Tellabs Tellabs 860 0 system in CDMA networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12 Tellabs Tel labs 8600 860 0 system functionality in a mobile network . . . . . . . . . . . . . . . . . . . . . . . . . . 13 The Tellabs Tellabs 860 0 system in wireline transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Network and service deployments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Managed migration path from the Tellabs 8100 8100 and Tellabs 6300 systems . . . . . . . . . . . 19 Network elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20 Element architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Network management system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26 . 26 Conclusions.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32 Conclusions . 32 Summary of product features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33 Acronyms and initialisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34

3


Network evolutio evolution n The Tellabs ® 8600 Managed Edge System has been designed using feedback from many global telecommunications service providers. Together they face the following market and technology challenges:  In

wireless networks, the 3G and release 5 ratification is accelerating the move of mobile services to an all-IP network; mobile service providers are therefore looking for a way to migrate their current Radio Access Networks (RANs) based on Asynchronous Transfer Mode (ATM) to a manageable Internet Protocol (IP) implementation.

 In

wireline networks, business and residential services are becoming predominately IP- and Ethernet-based; with increasing bandwidth requirements and new service delivery models, the current infrastructure infrastructure is becoming be coming inefficient and expensive to maintain.

 Fixed

and wireless convergence is taking place b oth in the services and in the underlying networks; the distinction between these previously distinct applications is now becoming blurred.

These challenges and the issues they raise are explained in more detail below. This is followed by a section explaining how the Tellabs 8600 system helps to address them.

Wireless Networks After years of speculation, mobile network operators are finally going through the evolution from 2G to 3G. This transition, which will call for major investments in the mobile network infrastructure, is taking place globally in both GSM- and CDMA-dominated markets. In practice, the evolution of mobile networks will mean upgrades to all mobile-network-specific components – from the mobile cell stations, through the RAN, and right into the core network. The underlying transport technologies are expected to undergo a transformation from Time Division Multiplexing (TDM) and Frame Relay (FR) to ATM and eventually IP. At the moment, the use of Internet Protocol / Multiprotocol Label Switching (IP/MPLS) (IP/MPLS) is restricted to the core of next-generation mobile networks but is gradually moving out into the RAN and towards the user. Eventually Eventually it will be used to provide an all-IP network, replacing the legacy ATM systems over the long term. Because the RAN infrastructure is expected to rely sooner or later rely heavily on IP, it makes economic sense to converge the fixed and mobile networks to use the same generic packet-based architecture. architectur e. As an example, many mobile operators intend to offer WiFi or WiMAX as complementary access methods in certain areas for different services. Additionally, business services such as the IP Virtual Private Network (IP VPN) or Metro Ethernet can be used to enhance the service portfolio.

Wireline Networks The increased utilization of the Internet and the move to IP-based applications and services is both a business and residential user phenomenon. One of the most popular business services is Local Area Network (LAN) interconnection, which is used to build corporate intranets and share company data and applications across remote sites. Traditionally raditionally,, the technology for providing LAN interconnection has been TDM, FR and ATM. However, However, these are all limited in their capacity and are not very cost-efficient when used to transport data 4

traffic in high volumes. The trend today is towards Ethernet- and IPbased services, which offer a more natural fit to the predominately data traffic mix. They are also more economical for the service provider to deploy, given the ubiquity of Ethernet-equi Ethernet-equipped pped network devices. This unified technology is blurring the boundary between the customer and service provider networks. It therefore offers service providers the opportunity to offer more value-added services such as VPNs, storage backup and outsourced IT applications. Now that bandwidth is becoming a commodity, users’ service expectations are rising. Simple connectivity as a service is no longer adequate; service providers need the ability to differentiate their services. Users are demanding Service Level Agreements (SLAs) from their service providers. These SLAs specify, for example, the service availability, repair time and servi ce quality parameters. To maintain these service levels, the service provider can also manage the customer premises equipment directly on the customer’s behalf. The main force for the growth of IP traffic has arisen from the everincreasing use of the Internet. This has been fueled recently by the now widespread availability of broadband services to home users. New applications continue to be layered on IP, IP, and legacy systems are already being replaced with IP alternatives – Voice over IP (VoIP) being a prime example. As broadband services become an essential part of everyday home life, the demand for bandwidth will continue to grow exponentially. This is creating pressure on the bandwidth available to the end user and is also creating scalability issues in the wider core network. The promise of triple-play services, where voice, data and video services are available through a single access interface, is another challenge for service providers to deliver cost-effectively.

The Wider Challenge An IP infrastructure is the ideal choice for quickly and costeffectively delivering different types of business and residential services. However, it is critical that it comply with the traditional requirements requiremen ts of the service provider provider:: carrier-class operations and manageability. For instance, wireless service providers need the Quality of Service (QoS), predictability and reliability that up to now could be delivered only with connection-oriented networks. The deployment of converged packet-based networks will enable wireless carriers to offer higher-value data services, which will enhance their existing voice services and create new revenue streams. With the Tellabs 8600 system, Tellabs offers service providers a scalable and potentially cost-effective solution to this challenge.

The Tellabs® 8600 Managed Edge System in next-generation networks The Tellabs 8600 system is a next-generation platform for building advanced telecommunications networks and services. It has been designed to meet the requirements of service providers who need to extend packet switching technologies more and more deeply into their access networks. While doing so, it provides the reassurance of a true carrier-class platform on which to build and deploy new services. Tellabs Te llabs understands that any investment made in the access

network has to last for many years. It has therefore designed a system that is so scalable and extendable that it is intended handle years of new service deployment and change. Tellabs also understands the pressures facing today’s ser vice providers. Falling margins in both fixed and mobile voice revenues, plus increasing regulatory and competitive pressure, are squeezing profit margins dramatically. Working with its global customer base, Tellabs has designed the Tellabs 8600 system to offer very low operation costs as well as rapid network and service deployment with an architecture that facilitates efficient use of the available network resources. Since existing Tellabs Tellabs customers have made significant investments in their current access infrastructure, Tellabs Tellabs provides a seamless transition and compatibility between its Tellabs ® 8100 Managed Access System, the Tellabs® 6300 Managed Transport System and the new Tellabs 8600 system. Full management capabilities across all of these platforms are provided by a single management system, the Tellabs ® 8000 Network Manager. The Tellabs Tellabs 860 0 system builds on the extensive experience Tellabs has gained with managed access platforms in over 300 deployments of fixed and mobile networks worldwide. The addition of IP/MPLS technology creates a robust, scalable and manageable platform for delivering next generation voice and data services. The combination of IP and MPLS provides the predictable properties of ATM but at the lower cost of Ethernet based devices. Extending MPLS into access and regional networks makes the entire network more controllable and efficient for transporting different types of technologies.

Figure 1. The Tellabs 8600 system’s position in service provider networks

As is shown in Figure 1, the Tellabs 8600 system is positioned in the access network to provide four basic applications:  Mobile

transport for 2G and 3G RAN

 Managed

IP VPN and Ethernet services

aggregation ag gregation for existing Te Tellabs llabs 8100 system and Tellabs 6300 system services

The key benefits of the Tellabs 8600 system, explained further in the chapters that follow, include provision of: A

platform supporting technologies needed for evolving mobile networks

 An

intelligent management system

 Design

for an optimized cost structure

Platform Supporting Techno Technologies logies Needed for Evolving Mobile Networks The Tellabs 8600 system when combined with the intelligent Tellabs 8000 manager is a solution for the needs of evolving mobile networks. It is a scalable platform that can be positioned anywhere in the mobile RAN. At the Radio Network Controller (RNC) site it can provide significant savings for the service provider in the overall transport cost by improving scalability and optimizing the RNC port costs. Additionally at the RNC site, the Tellabs 8600 system platform can act as a Customer Edge (CE) network device. Positioned at the hub site, the Tellabs 8600 system can bring significant CAPEX and OPEX savings by optimizing the bandwidth and improving the management of the transport network through its testing capabilities and protection solutions. When positioned at the base station site, the Tellabs 8600 system platform can aggregate different protocols, encapsulate them into MPLS Pseudo Wires and statistically multiplex them over various types of backhaul links. Mobile networks are evolving, and significant changes need to be made to enable new high-speed data services in the mobile RAN. In a single platform, the Tellabs 8600 system supports the technologies needed in mobile transport transport network evolution. It can handle the transition in moving from 3G R99 ATM networks to 3G R5 IP/MPLS and Ethernet networks in a cost-effective manner. At the same time, the Tellabs 8600 system solution can still also carry 2G TDM traffic, providing a single platform for mobile network transmission. The Tellabs 8600 system platform optimizes network capacity by using sophisticated Quality of Service and Traffic Engineering (TE) tools to support the growth of mobile data services. By using standards-based signaling and network control mechanisms between the network elements, it is possible to reserve explicit paths for time-critical, time-critic al, delay-sensitive or bandwidth-intensive connections through the network. Less time-critical or bandwidthintensive connections can be allocated along shared paths, where bandwidth and delay parameters are more flexibly defined. Wirespeed forwarding and full resiliency mechanisms are the keys to very high-speed, predictable performance. The Tellabs 8600 system is designed for fixed and mobile network convergence. In addition to the Ethernet connectivity, it can support any combination of mobile and fixed backhauling, such as E1, ATM IMA, POS and channelized STM-1 interfaces, in the same network element.

 Multiservice

These applications are described further in the following chapters, which outline the role of the Tellabs 8600 system in wireless and wireline networks. In many of these applications, the Tellabs 8100 system and the Tellabs 6300 system solutions currently play an important role. These are expected to continue to remain in service provider networks for many years to come. Integration between these Tellabs platforms is covered l ater in this document.

Intelligent Management System Because the number of network elements in the access and regional network is an order of magnitude larger than that in the core network, effective network management is absolutely essential. This imposes additional scalability requireme requirements nts for the management system.

5


The Tellabs ® 8000 Network Manager is an integral part of the whole solution. It provides a full set of tools based on an easy-touse graphical user interface to manage element-, network- and service-level configurations. Since the operational management expenses can be as high as 80% of the overall costs of the network, this is a key feature of the overall solution. The Tellabs 8000 manager offers significant advantages for the service provisioning process and management of large networks. Traditionally raditionally,, service ser vice provisioning has been performed per formed using element management systems or even industry-standard command-line-based tools, which is often a complex and timeconsuming process. Management complexity becomes much greater when the network grows and several technologies are involved. According to some service provider statistics, these tools can result in a first-time success rate of just 60% for provisioning of individual network elements. This low success rate leads to significant costs and increased lead times in delivery of new services. With the Tellabs 8000 manager, service provisioning is a highly automated process, with the system taking care of creating all of the parameters and configuring the relevant network elements. In the same way, making changes to services or network connectivity is very straightforward and quick. Each connection or service can even be individually tested before launch or even while it is operational. For service assurance, the operator can see how network or element interruptions are affecting individual services, enabling much faster reaction to changes. Most importantly, with the Tellabs 8000 manager manager,, monitoring accuracy and management capabilities are not sacrificed, even when the network scales to tens of thousands of elements. This can give a huge competitive edge to a mobile operator with a network facing heavy growth. The Tellabs Tellabs manager also operates with open interfaces enabling data to be retrieved or sent to other Operational Support Systems (OSS) that are deployed in the s ervice provider environment. All network- and service-related information is stored in a database, which is accessible using open Application Program Interface (API) standards.

Design for an Optimized Cost Structure The network elements of the Tellabs 8600 system solution vary in size, which facilitates the best fit for every network location. The platform is flexible, for various applications, and therefore it can serve essentially all of the transport requirements in the access or regional network. The modular design of the Tellabs 8600 system platform provides the flexibility to equip each element with different capacities and interfaces as required. These are specified according to the service and network requirements, which typically vary with the position in the network hierarchy. Depending on the changing requirements requireme nts of the data communications infrastructure, infrastructure, the network capacity and interfaces can be adjusted and upgraded throughout the service life cycle. With traditional network element solutions, the cost of separate switching cards for each platform can become very significant when one considers the overall cost of the network element. This is especially true where some network elements are supporting only a few customer interfaces. This can make the initial network deployment expensive for new services that may start off with low customer volumes but require the deployment of many net work elements to reach the target customer market. With the Tellabs 8600 system, it is profitable build out the network 6

even with a small initial quantity of bandwidth and services. This is because the cost of service entry is significantly lower than with a traditional network elements, which can lead to reduced payback times and a quicker return return on investment. With W ith a distributed switching architecture, switching capacity is increased as new interface cards are added. This reduces the service entry cost, since the basic configuration is very simple, even with full elementand network-level redundancy features. Expanding the network to support a larger number of services is achieved by simply adding new interface cards to the platform. Increasing the scope of the network to cover denser and larger geographical areas is as easy as installing new network elements and follows the same “pay as you grow” principle.

The Tellabs 8600 system’s benefits and role in wireless transport After years of speculation as to whether 3G evolution will ever happen, network deployments have finally started and many operators have already launched or are currently in the process of deploying 3G services. The most important 3G standards are UMTS and CDMA2000. UMTS networks use WCDMA radio technology, and they are often referred to accordingly as WCDMA networks. This section of the document discusses how the Tellabs 8600 system can provide a solution for these networks. When determining the optimal solution and technology for a 3G network, the service provider should consider at least the following issues:  The

investment being made is for the long term, and the network should be able to scale easily in the future.

 The

solution must take into account the need for service and network convergence, where multiple types of service can be offered on the same platform.

 The

network management solution must support the business processes and be able to lower the operational expenses significantly.

When compared to the alternatives, the Tellabs 8600 system offers an attractive and potentially long-term solution to the mobile RAN transport challenge. Typical Typical RAN transport solutions on the market today are based on TDM or ATM technologies. These are old technologies that have limited capacity and do not provide the long term scalability and flexibility that mobile network evolution demands. As IP eventually becomes the native transport protocol throughout the mobile network, new data-rich services will drive the need for higher-bandwidth services to emerge. Also, with the need for fixed and mobile network consolidation to reduce operating costs, service delivery using Ethernet and MPLS will become the norm. Legacy technologies will not be able to support the unstoppable move to converged network architectur architectures. es. The most significant benefits the Tellabs 8600 system solution brings to mobile RAN networks can be summarized as follows: A

single-platform solution for 2G/ 3G architectures and beyond

A

potentially long term investment with IP/MPLS support from day one

A A

single management platform for all Tellabs mobile solutions

highly integrated architecture with carrier-class operations and low inventory cost

Single-platform Solution for 2G and 3G In the move from 2G to 3G and beyond, the transport technology moves from TDM and FR to ATM and eventually to IP. The use of dedicated solutions for different transport needs is costly from both a CAPEX and OPEX point of view. The Tellabs 8600 system can handle the aggregation and transport of all of these protocols in parallel. This makes perfect sense at, for instance, new or existing sites where 2G and 3G are collocated. The Tellabs 8600 system is a vendor independent transport solution that c an interoperate with other vendor’s technology components in the BSS and RAN infrastructure. And, most importantly, the transport network can be managed even as a single entity.

Otherwise, with both an ATM and an IP platform to look after, management and maintenance maintenance costs will remain high. The amount of savings gained in terms of leased line rental is a highly marketdependent figure and in certain markets can be huge. Once the Tellabs 8600 system solution is deployed close to the Node-Bs in the access network, the transport infrastructure can be further optimized with more cost efficient Ethernet interfaces. Also, the available bandwidth can be utilized in a more efficient way by allowing overbooking for data services. Ethernet interfaces can be used to enable new Metro Ethernet services and Ethernet leased lines for backhaul. This use of Ethernet devices can further lower the total cost of the RAN.

Long Term Investment

A Single Management Platform

Today’s mobile transport solutions s olutions are based on TDM and ATM technologies. These are optimal for 2G and for the first releases in the 3G UMTS networks. However, the longer-term evolution is expected to bring IP all the way to the access network and even out to the mobile terminals. At the same time, with the introduction of the High-Speed Downlink Packet Access (HSDPA) and CDMA 1xEvolution – Data Only (EV-DO) high-speed data services, the bandwidth capacities will increase significantly. This will pose new challenges for the access and transport networks. The TDM- and ATM-based network infrastructure will cease to be cost-efficient or even capable of meeting these challenges, especially since these technologies do not figure significantly in most telecommunications equipment vendors’ product strategies. The Tellabs 8600 system is already based on IP/MPLS, which does not require leapfrog investments or forklift upgrades in moving to the all-IP phase. Additionally, it allows the operator to offer new wireline services, such as Ethernet and IP VPNs, on the same platform.

Tellabs has over 15 years of experience in developing power ful Tellabs service providers’ network management tools in cooperation with our customers. Network management has always been an integral part of the Tellabs access platforms and has proven to be a key differentiator in the market. With hundreds of networks based on the Tellabs solution, Tellabs Tellabs is now making network evolution and transition to new technologies even simpler. With a single network management system, the service provider can manage the whole network. Also, the service provider can deploy the solution without cost-intensive integration work and with minimal investments in the existing management platform. The ability to use the same personnel and processes without major retraining makes the change extremely straightforward.

Figure 2. 3G RAN solution with the Tellabs 8600 system solution

Figure 2 illustrates the comparison of business cases for an ATM and a Tellabs 8600 system IP/MPLS-based RAN solution. The main cost savings in favor of the Tellabs 8600 system solution come from the reduction in CAPEX, since there is no need to invest in additional IP routing functionality at each site when moving to 3G R5 and beyond.

Integrated Architecture with Carrier-class Operations The Tellabs 8600 system was initially specified, and has been continuously developed in partnership, with our service provider customers. All hardware elements and the network management system are built with high reliability, performance, scalability and cost-efficiency in mind. The same highly integrated hardware and software architecture is used across the whole platform. The distributed switching architecture architecture is the main factor that can make the system cost efficient to deploy even at small sites. The basic configuration of the element is kept very simple yet retains the flexibility to equip each element with the correct mix of interfaces. This makes it suitable for very different locations and applications. A broad range of element- and network-level protection options can be instituted on the basis of the network availability requirements.

The Tellabs 8600 system in GSM and UMTS networks Currently, mobile operators are required to build their RAN Currently, infrastructure according to a system that was specified long before MPLS was considered mature enough to win acceptance in service provider networks. Today, Today, UMTS Release 99 (R99) is the only application that requires ATM backhaul backhaul for efficient ef ficient transport in the service provider networks. However, WCDMA implementations deployed in the U.S. and Asian markets are already IP-based.

Based on business cases prepared by Tellabs, these savings can amount to approximately 80% of the cost of the equivalent multipleplatform network. Approximately 20% savings in OPEX can result from the reduction in leased line costs due to the simpler transport infrastructure infrastructur e required by the more scalable Tellabs Tellabs 8600 86 00 system. sys tem. 7


The strict R99 ATM requirements do not go away with the introduction of MPLS. But, by limiting ATM to the edge of the RAN, the required level of efficiency and robustness can be achieved with the deployment of a more modern access network. With a Tellabs 8600 system solution, the RAN network can simultaneously handle the access for any base station through each of the evolution phases as illustrated in Figure 3.

Tellabs 8600 system network. In a similar way, dedicated paths with specific priorities can be provisioned for any type of connection across the Tellabs 8600 system network. Figure 4 below shows where the different Tellabs 8600 system network elements can be positioned in the mobile RAN.

Figure 3. Multiple networks v. IP/MPLS in 2G and 3G RAN

Over time, the RAN transport is expected to change to IP. This change applies from R5 onwards. All communication in the mobile network should eventually be based on IP, IP, and mobile terminals identified with IP addresses. The Session Initiation Protocol (SIP) will be the method used to set up and tear down such connections in the mobile network. With the Tellabs 8600 system, the service provider can plan this network evolution so that there is no need to change the transport infrastructure even though the network includes different locations and different stages of evolution. With the introduction of the IP Multimedia Subsystem (IMS), (IMS), the whole network and all devices will be using the same services, independentt of their access technology. True service convergence independen will be enabled. Some of the leading service providers have already begun implementing an IMS infrastructure. The advancing network and service evolution also brings new data services, with higher bandwidth and quality management needs. This poses new challenges for the network transport as well. No single technology will meet all of the requirements in terms of cost, scalability or flexibility. The Tellabs 8600 system solution covers the transport part of the mobile access network from the base station sites to the RNC/BSC sites. The launch of 3G networks is driving the need to build a new and scalable transport infrastructure for these services. In particular,, the emergence of HSDPA and High-Speed Uplink Packet particular Access (HSUPA) services will dramatically increase the bandwidth requirements from the cell sites. Even though 2G networks are already built out in the most developed markets, in certain areas there might be locations where there is a need for adding more GSM base stations. In these cases, the use of one transport solution for each site is an advantage. The Tellabs 8600 system can be used for TDM transport in the same way that it is used for ATM.

Figure 4. Tellabs 8600 system positioning in 2G/3G RAN

RNC Sites Outside the network core, the first optimal position for a Tellabs 8600 system platform is at the RNC site. This network element is typically a fully redundant and highly scalable Te Tellabs llabs ® 8660 Edge Switch. It can also be used to connect RNC sites together in the mobile core network. In addition to 3G traffic, the same network element can be used to aggregate traffic from the 2G GSM network into the BSC. This can handle the case where the 2G network is being upgraded with new equipment or additional sites and would avoid the need for investments in a separate infrastructure including including network elements and management systems. With 3G traffic aggregation the Tellabs 8600 system solution is significantly more economical and scalable than traditional ATM switches. For R9 9 applications alone with E1 interfaces and IMA towards the Node-Bs and STM-1 ATM ATM handoff towards the RNC, the operator can save s ave 50% per E1 (see Figure 5). Additionally, Additionally, the Tellabs 8600 system platform today offers direct and potentially extremely cost effective Ethernet interfaces and routing capabilities, which become essential at least with the future deployments.

In practice, both the ATM and TDM traffic is carried through tunnels that are provisioned with predefined c apacities through the

Figure 5. Tellabs 8660 switch at RNC site

8

Using the Tellabs 8600 system at the RNC site offers the service provider the following potential advantages:  It

improves the network scalability and allows more Node-B sites to be controlled from one centralized RNC.

 It

reduces CAPEX by reducing the number of interface ports needed on the RNC. This is as a result of performing the ATM inverse multiplexing and VP/VC grooming on the Tellabs 8600 system platform.

 It

Figure 6. Tellabs 8600 system at hub site

 It

In summary, use of the Tellabs 8600 system at the hub sites can bring the following business benefits:

 The

 It

enables the use of lower-cost unchannelized interfaces at the RNC site. can further reduce CAPEX since the same network element can be used as part of the mobile core. same platform used for 3G traffic aggregation can be used for grooming 2G GSM traffic arriving on TDM links.

allows a smooth network migration from TDM to ATM transport and eventually to IP.

 Bandwidth

Hub Sites Positioning the Tellabs 8600 system platform closer to mobile base stations can yield additional business benefits. These include better bandwidth utilization, more options for backhaul technologies and improved network management capabilities. The role of the hub site is to aggregate different traffic streams, including voice and data, from the access network into the mobile core network over fewer connections. The introduction of a Tellabs-8600-like MPLS network infrastructure can significantly optimize the bandwidth utilization, enable use of cost-efficient Ethernet interfaces and reduce the number of leased lines required to carry the traffic. It also allows the traffic to be handled with finer granularity. Depending on the bandwidth, port density and redundancy redundancy requirements, the hub site can be implemented using the Tellabs ® 8660 Edge Switch, the Tellabs® 8630 Access Switch or the Tellabs ® 8620 Access Switch. The business case for utilizing the Tellabs 8600 system solution at both RNC and hub sites is persuasive, and this solution can generate significant savings in both CAPEX and OPEX terms. Use of the Tellabs 8600 system solution for the hub sites saves on bandwidth costs, not only because of the number of lines required but also due to the ability to move from Constant Bit Rate (CBR) to Variable Bit Rate (VBR) transport. This is the gain from statistical multiplexing, which makes sense with increasing and bursty data traffic. From some business case calculations with our customers, we have determined that distributing only one hub layer to the network can yield more than 25% cost savings in E1 leased line costs. Naturally, if alternative, more cost efficient transport technologies are used, the percentage is higher. The Tellabs Tellabs 860 0 system is flexible in this sense and offers various alternatives, such as Ethernet connectivity, which is becoming more and more attractive (see Figure 6). It should be remarked that in a converged network, customers can be connected to other services over the same Tellabs 8600 system platform reaching the hub sites. In fact, the service can even be implemented with the Tellabs 8600 system and related management system. Both of these improve the operator’s business case and service manageability.

utilization is improved through traffic grooming and network overbooking.

 The

solution is scalable for higher-bandwidth data services such as HSDPA.

 ATM

traffic can be monitored and tested over the whole connection. This is particularly important when statistical gain is applied.

 Low-cost

Ethernet interfaces interface s can be used for implementing Ethernet leased-line transport.

 Additional

value-added services can be carried c arried and managed on the same network infrastructure. These can include managed IP VPN and Ethernet services with differentiated SLAs.

Base Station Sites When the Tellabs 8600 system platform is used at the base station sites, traffic can be consolidated onto a single access network infrastructure, which brings savings in transport costs even in the local loop. Additional savings are gained through statistical multiplexing, which is significant when higher-speed data services are brought into use. It is worth noting that, although there is only one physical connection, the different traffic streams can be managed logically as independent connections with their individual service quality requirements. Locating the Tellabs 8600 system unit at the cell site, like in Figure 7, allows the access network to be managed end to end so that modifications – for instance, to the capacity or service quality settings – can be handled remotely. When the Tellabs 8600 system is deployed to the cell site, there are several alternative ways to arrange the access to the network. Depending on the open infrastructure and service requiremen requirements, ts, the service provider can choose Ethernet, DSL or TDM. PDH and SDH are the most popular alternatives today, since this infrastructure is widely deployed and reasonable for voice transport. In the beginning, the 3G traffic is indeed mainly voice. However, when the HSDPA-based services are launched, capacity requirements grow and other alternatives are likely to be worthy of consideration. From a cost and availability point of view, DSL service is attractive for connectivity. Because the base station can separate voice and data traffic, voice can be directed to the TDM transport system and data to DSL via Ethernet. This seems to be a cost-effective set-up. For instance, DSL’s cost advantage over E1 when backhauling HSDPA data traffic in the local loop is about 50%. In the context of RAN’s transport as a whole, even 59% cost savings can be achieved over ATM-based RNC application. 9


Another inviting option is to utilize the Metro Ethernet networks, which are already widely deployed. Also, the bandwidth cost is typically significantly lower than that of traditional TDM networks. In using Ethernet transport in the RAN, it is important to make sure that the synchronization and service quality management can be arranged properly. properly. With Tel Tellabs labs 860 0 system functionality functionality,, these important elements are well supported. Synchronization is discussed later in this document.

links and rebuilds all of them at the selected interfaces of another element. This not only makes the process fast but also facilitates correct configurations for all of the related network elements. In this context, the operator can even verify the connectivity remotely with the management system tools. The Tellabs 8000 manager can offer the service provider the following advantages:  Fast

response to network net work changes with remote configuring, automated provisioning and testing

A

single management solution for multiple access technologies, including TDM, ATM, FR, IP and Ethernet

 A

carrier-class network manager built on the basis of service provider needs, supporting 30,000 network elements and tens of concurrent users

 An

easy-to-le arn and -use management system with a graphical easy-to-learn user interface that hides the network complexity from the user

 Potentially

significant cost savings for operators through provision of management for multiple technologies, remote management and fast troubleshooting

Figure 7. Tellabs 8600 system at cell site

The Tellabs 8600 system is very flexible, allowing the service provider to use any of the mentioned technologies for access. Naturally, it gives the service provider the choice of using the platform for some base station sites while using traditional SDH platforms for the last-mile access for other sites. This, of course, depends on the bandwidth and ser vice requirements, plus the growth expectations for each area. For a base station access solution, the Tellabs 8600 system provides the following main benefits:  It

can aggregate different 2G- and 3G-related protocols and traffic streams on the same platform.

 Connection

and service parameters can be changed remotely via the network management system.

 New

services can be implemented on the platform to attract new customers and increase revenue streams.

 Cost-efficient

and scalable Ethernet links can be used for backhauling the traffic into the RAN.

Efficient Management in Mobile Networks Mobile networks are by definition very dynamic in nature and are growing especially dramatically now. New base stations are often constructed, and bandwidth links are frequently upgraded or added. When the network is first built, the sooner it can be put into service, the sooner the service provider can turn on its revenue stream. Network management has a critical role in all of these processes. The sophisticated tools of the Tellabs ® 8000 Network Manager provide major benefits for the service provider throughout the network life cycle. They support day-to-day operations throughout the continuous evolution of the network, providing endto-end connectivity management. Managed re-hosting capability is one excellent example of the Tellabs 8000 manager’s competencies. When a selected group of connections must be moved from one location to another, the operator can with one command execute the whole operation. The Tellabs 8000 manager automatically tears down all of the selected

10

Network Convergence The boundaries between fixed and mobile services and networks are vanishing. Deregulation is opening up new opportunities for service providers. The resulting competition is driving every service provider to extend its service portfolios. Ideally, the same network infrastructur infrastructure e and the same s ame management system should be capable of handling all of these different services. The Tellabs 8600 system is designed exactly for this purpose and uses MPLS for convergence, as shown in Figure 8. Convergence can be executed at various levels and depends greatly on the organization boundaries. One way to segregate the various levels of convergence is:  Mobile

convergence in terms of providing 2G and 3G ser vices with the same platform

 Fixed

and mobile convergence where the service provider not only offers mobile services but also, e.g., produces broadband or business services, or just transport from a unified infrastructure

 Service

convergence, of which the IMS infrastructure and the same service offering independent of the end-user device is a good example

The access network is the most expensive part of the service provider’s overall infrastructure. infrastructure. Therefore, the use of common multiservice-capable elements and flexible management tools in the access network offers the best potential savings for service providers.

The Tellabs 8630 switches were deployed to the aggregation sites to allow utilization of a single platform for both 3G traffic and residential DSL services. This lowered the CAPEX related to the use of several devices at the hub sites and optimized the costs related to the backhauling. With the help of the Tellabs 8630 switches, operators were able to utilize cost-effective Ethernet backhauling for RAN and solve synchronization challenges often related to ME backhauling. The Tellabs 8600 system solution offered superior QoS features and end to end resiliency as required in large-scale Metro Ethernet backhauling cases. There are already hundreds of Tellabs 8600 systems deployed in this network.

Customer Case with RAN Optimization

Figure 8. Service convergence enabled by MPLS

If we look at network convergence from a mobile operator’s point of view, the following types of services and access technologies are of interest:  WiFi

and WiMAX access used as complementary wireless access technologies

 Ethernet

and IP VPN services s ervices for business customers

 Broadband  Wholesale  Tr Transport ansport

Internet access for residential customers

bandwidth to offer to other service providers

for Data Communications Networks Net works (DCNs)

The technology evolution from ATM to Ethernet and IP is taking place everywhere, not only in mobile networks. For instance, DSLAMs, which are the primary method for implementing broadband Internet access services, are moving from using ATM to Ethernet for their backhaul protocol. The combination of multiservice interfaces and MPLS PWE tunnelin tunneling g on the Tel Tellabs labs 8600 system platform makes the evolution path easier for service providers.

In this RAN optimization reference case, the operator was launch launching ing 3G services with a tight schedule due to the fierce competition in the market – at the same time, two other operators were also launching launchi ng 3G ser vices. The main requirements for the first-phase implementation were:  Rapid

deployment of 3G transport to allow rollout of the first 3G services

 A

scalable platform that can support future growth of the services and replacement of existing ATM devices that had been implemented for 3G test sites

 Full

R5 compatibility from the first installation, to minimize the cost of the transport network and eliminate future forklift upgrades

 In

terms of network management a smooth migration from the existing network to the new infrastructure

The proposed solution to meet these requirements was a Tellabs 8660 system collocated with the RNCs to:  Optimize

RNC port costs – utilization of unchannelized interfaces towards the RNC

 Enhance

the scalability of RNCs and RNC front nodes – a single device with high-density interfaces and support for all requirementt in a single node requiremen

 Ease

Customer cases Before providing a description of the Tellabs 8600 system’s roles in the mobile backhaul, we consider two examples of how and why this solution was chosen for specific networks. These two cases are chosen because they differ from each other in their drivers and requirements. This shows the flexibility and uniqueness of the Tellabs 8600 system solution.

management and connection creation from the very beginning of the commercial 3G solution

The access transport part of the network relied on the existing SDH-based transport network and external leased lines. This allowed rapid launch of the service because only new elements were located at the RNC sites. At this stage of the implementation, utilization of existing platforms was seen as the most cost effective solution.

Customer Case for Building a Converged Network This operator was looking at transport solution that lowers the capital and operational expenditures when moving to a converged network. The objectives for the transport network project were: A

single converged network to operate and manage all services

 Utilization

of Metro Ethernet backhauli backhauling ng to lower the cost of

transport A

long-term solution with R5 support

 Halting

of investment in ATM platforms

 Finding

of a cost-effective solution for multiservice aggregation

sites 11


It was clear to the operator already in the initial planning of the commercial 3G deployment that the transport network had to be capable of accommodating growth in the capacity, number of Node-Bs and proportion of data in the network. The second phase of the network implementation was planned to support HSDPA launch.. To support the increasing amount of data traffic in the launch mobile backhauling, a second layer of MPLS-aware devices was implemented for the network. The new sites – aggregation sites – were designed to address the following issues:  Adding

statistical gain to the data traffic and offloading the ATM at the aggregation point to minimize capacity needs and costs of backhauling

 Optimizing

the utilization of leased connections between cell sites

and RNCs  Enabling

low-cost transport alternatives like Metro Ethernet backhaul and EoSDH transport

 Simplifying

network building and modifications (one touch reparenting) with advanced management solutions customized for mobile access

 Obtaining

visibility of leased leas ed line quality and enabling testing in the access part of the network

 Offering

a common infrastructure for 2G and 3G already at the aggregation point

 Offering

the possibility to support additional services like Ethernetbased WiFi and WiMAX in transport but also IP-based services for several network locations

The transport between aggregation points is still leveraging the existing SDH network in part – now in Ethernet over SDH mode. Ethernet interfaces are used towards the SDH network to optimize the spares management and to equip the network for the future. As capacities grow even higher, Ethernet interfaces allow rapid upgrades to Ethernet leased lines, Metro Ethernet backhauling or utilization of direct fiber links wi th GE. Currently, there are approximately 40 Tellabs 8600 systems deployed in this network.

Figure 9. Overview of the CDMA network

exceeds that, there are multiple, parallel links. This me ans that the IP traffic needs to run over ML ML-PPP -PPP.. Ethernet connectivity is another alternative to multiple TDM links. Some base stations already offer an Ethernet interface, and many vendors have this on their roadmaps. Unlike WCDMA, ATM technology is not present at all in the CDMA access. Figure 9 shows a typical mobile operator’s network with 2G and 3G components. It defines the basic building blocks in the CDMA network as well as the connectivity in the access network. TDM – more specifically, PDH and SONET –still dominates the access. Particularly in the U.S., a typical mobile operator leases all of its transport from another operator and owns only the mobile service specific parts of the network. Conversely, in Europe the mobile operators tend to invest in at least some part of the access transport and they more often utilize microwave links instead of fixed lines.

The Tellabs 8600 system in CDMA networks Another common way to implement 3G networks is using a CDMA2000 technology path, which is especially popular among a number of U.S. operators but also is deployed in certain countries in Asia and Latin America. It should be noted that in the U.S. some operators have chosen a GSM and WCDMA path to follow instead of CDMA. Transition to 3G has been particularly strong in the U.S. and Asia Pacific region. The air interface is naturally different from WCDMA. From the transport point of view, the main difference between WCDMA and CDMA is the protocol carrying the traffic. Most of the operators have started their transition to 3G with 1xRTT technology, which could be considered to be a 2.5G phase, and have now initiated the rollout of the 3G network with EV-DO. However,, some operators have announced a move to EV-DV directly However from 1xRTT. With 1xRTT, all of the traffic is based on FR, whereas with EV-Dx the traffic from the cell site is IP and carried over PPP or HDLC. The first step when EV-DO is deployed is to carry only data over IP, while voice remains in FR (1xRTT). Only with EV-DV is all traffic IP-based, but that phase remains for the future. Because the physical connectivity toward a cell site today typically consists of E1 or T1 links and the total capacity requirement

12

Figure 10. Tellabs transport solution for CDMA networks

The role of the Tellabs 8600 system is similar to what was described for W CDMA. In other words, it covers the mobile transport network from cell site to the BSC as shown in Figure 10. The driver for using the Tellabs 8600 system platform for traffic aggregation is mainly to minimize the operational costs relating to the cost of bandwidth. This becomes more and more essential with the growth of data traffic and the increasing capacity.

Potential benefits provided are:  M Minimized inimized

cost of bandwidth

 One

platform for various traffic needs and services (costefficiency in terms of investments and maintenance)

 Improved

management, fast response to net work growth and ease of topology changes

 Readiness A

tunnel is terminated at the edge of the MPLS network domain, where the label is removed and the TDM traffic is passed to the destination element or out into the TDM network. This process is illustrated in Figure Figure 12. All TDM traffic is carried transparently transparently through the MPLS domain, and bandwidth can be reserved for each LSP that the Pseudo Wires traverse.

for convergence and flexibility for various technologies

cost-ef ficient solution cost-efficient

Tellabs 8600 system functionality in a mobile network Multiprotocol Grooming and Transport MPLS technology is an ideal transport solution for evolving mobile networks. It can handle all of the protocols required in each of the 3G release phases. MPLS can carry traffic over any underlying transport network. Any Layer 1 or Layer 2 protocol can be transparently transported over the MPLS network using Pseudo Wires (PW). These are sometimes referred to more specifically as Pseudo Wire Encapsulation Edge to Edge (PWE3). The PW connections can be regarded as permanent connections just like ATM PVCs. Each PW connection can reserve an explicit amount of bandwidth from the network and can be protected end to end through the network if required required.. The Tellabs 8600 system platform can combine the functionality of a number of network elements. For GSM and UMTS traffic aggregation, the most important facilities that the Tellabs 8600 system provides are TDM and ATM cross-connections as well as IP routing on a single device. With CDMA, instead, FR and PPP or HDLC are essential from the transport point of view. The multiprotocol connectivity available is shown in Figure 11. 11.

Figure 11. Multiprotocol connectivity through the Tellabs 8660 switch network element

TDM is the most common access technology used in GSM networks. Traffic to and from 2G base stations goes over channelized channel ized E1 links or STM-1 links in the SDH network. Instead of using a traditional TDM cross-connect for this task, the Tellabs 8600 system platform can be used as the first aggregation element in the mobile access network. It combines the TDM cross-connection functionality with ATM switching and IP routing. Cros s-connections or traffic grooming can be performed at the timeslot level (DS0). Traffic can be switched between channelized interfaces as in traditional TDM cross-connects or towards an MPLS interface on the same platform. At the MPLS interface, the TDM traffic is encapsulated by adding an MPLS label and sent through the PWE3 tunnel over a Label Switched Path (LSP). The other end of the

Figure 12. ATM and TDM cross-connections, and transport between Tellabs 8600 system elements

This technique makes sense when the service provider is focusing its investment on long-term transport solutions and wants to optimize the infrastructure to lower the total cost of ownership for the network. Instead of using a separate platform for each type of transport needed, a single Tellabs 8600 system solution with one management system can fulfill all of the mobile transport requirements. For UMTS networks, the traffic from a Node-B is currently transported over ATM. ATM VP/VC circuits can be cross-connected just like like TDM TDM timeslots. timeslots. When When these connec connections tions are are made made from one ATM interface to another, the element looks ex ternally like an ATM switch. Through implementation of a Tellabs 8600 system solution at the base station site, ATM connections can be carried over MPLS Pseudo Wires transparently. These Pseudo Wires are transported along MPLS LSPs, which can be assigned a traffic class according to the ATM Class of Service. The Tellabs 8600 system platform also supports ATM IMA functionality in all of its channelized channel ized interfaces. This means that an ATM IMA group coming from a Node-B can be terminated at the Tellabs 8600 system element. More cost-efficient interfaces and transport mechanisms can then be used in the transport network. Typically, the physical link to the cell site is E1 or channelized STM-1. STM-1. Over the longer term and from R5 onwards, the ATM transport will be replaced with IP. When UMTS R5 is deployed, the RAN starts to migrate to a fully IP-based network. The Tellabs 8600 system elements are essentially IP routers with MPLS support for all interface types. In the R5 specification, the physical connectivity to the Node-B is either a channelized TDM or Fast Ethernet. If the connectivity is based on multiple E1 links using Multilink PPP (ML-PPP), these can be terminated in the Tellabs 8600 system element in a similar way to ATM IMA. Frame-Relay- or HDLC based traffic, which is often present in CDMA networks, can be transported via Pseudo Wires. The Pseudo Wire connections are independen independentt of the protocol transported and can be provisioned end to end with the Tellabs 8000 manager’s easy-to-use graphical tools. Before going live,

13


Traf ﬁc class

Conversational RT

Streaming RT

Interactive best effor t

Background best ef for t

Fund Fu ndam amen enta tall chara charast ster eris isti tics cs

Prese Pr eserve rve time time rela relati tion on (var (varia iati tion on)) between information entities of the stream. conversational pattern (stringent and low delay).

Preserve time relation (variation) between information entities of the stream.

Request response pattern. Preserve payload content.

Destination is not expecting the data within a certain time. Preserve payload content.

E xample of the application

Voice

Streaming video

Web browsing

Background download of emails

ATM Ser vice Categor y

CBR

r t-VBR

UBR+

UBR

DiffSer v Traffic Class

EF

AF1(*)

AF4(*)

BE

(*) Use of AF traffic classes is operator dependent

Table 1. Service classes according to 3GPP TS 23.107

the connections can be tested to ensure that they deliver the desired functionality. The service provider can also monitor each connection in the Tellabs 8600 system network in real time and get statistics and faults mapped to individual connections. This helps the service provider to understand, for example, the impact that a network fault can have on specific connections or services.

Service Quality Management

Synchronization Management

Current mobile services are predominately voice-based, a situation that is likely to prevail until the beginning of UMTS deployments. However, new types of data and multimedia services will become more and more popular. This mixture of voice and data services will set new service quality requirements for the network. It will need to be able to handle these requirements in an efficient and appropriate manner.

Synchronization plays an important role in mobile networks since the base stations must be well synchronized to ensure good voice quality and manage the call hand-overs.

The UMTS specifications define four service classes, which are listed in Table 1. Each service within a given class has a common set of characteristics.

GSM and WCDMA networks typically obtain synchronization with the cell site from the E1 or T1 leased line or the microwave link to which they are connected. When the connectivity is TDM, synchronization is not an issue. However However,, where Ethernet connectivity is concerned, timing could become problematic. Traditional Ethernet networks do not have the ability to provide a clock-based signal to a cell site. Standardization bodies are currently working with this issue, and some candidates are already present. These are IEEE’s 1588 Precision Time Protocol (PTP) and Synchronous Ethernet. IEEE 1588 was originally specified for Local Area Networks for use with testing. The second version, which adds support for the WAN environment, is still in progress. Synchronous Ethernet is described in ITU-T G.8261, which specifies the method of distributing the synchronization via the Ethernet line signal. Tellabs follows closely the standardization progress and has implemented the Synchronous Ethernet. Ethernet. PTP is intended to be implemented soon after the specification exists. With the Tellabs 8600 system, the synchronization can also be relayed to the cell site by means of adaptive timing, where a TDM interface in the Tellabs 8600 system element can obtain synchronization through a TDM Pseudo Wire. It is worth mentioning that Tellabs 8600 system elements are, in fact, part of the synchronization network so it can distribute the clock to other elements in the network.

The transport network must be able to implement these service classes in the appropriate way throughout the whole network. They can be supported using any of various transport technologies or even with a combination of them.

With CDMA networks, the synchronization and packet network issue does not arise from the transport point of view since CDMA uses GPS receivers at each cell site. This is, naturally, an option also with WCDMA networks, but it is not widely deployed. More often, since the 2G and 3G base stations are collocated and SDH is present as well, one could obtain the synchronization through SDH. As described for the service provider, Tellabs offers various options for arrangement of the synchronization in the net work and therefore removes the barriers from migration to packet-based backhauling.

14

The Tellabs 8600 system has extensive support for traffic quality management. Traffic Traffic forwarding for warding inside the network element is performed at the hardware level to facilitate wire-speed performance for all traffic. For high-priority traffic, bandwidth can be reserved through the network using the RSVP-T RSVP-TE E signaling and connection admission control (CAC) protocols in the elements along the signaled path. These protocols facilitate that the requested connection can be established without disturbing the existing traffic and that the reserved path is always available for this connection. Tunnels run over MPLS LSPs, which can be configured to carry traffic for one QoS class or for a mixture of QoS classes. Each LSP can be configured with different parameters for path protection and bandwidth reservation depending on the type of traffic it is carrying. The Tellabs 8600 system implements QoS management using IP DiffServ and maps other protocols to the DiffServ traffic classes to provide end to end service quality. For ATM service classes, the Tellabs 8600 system platform supports CBR, VBR, UBR+ and UBR service categories. Traffic forwarding, queuing, queuing, scheduling and shaping is performed on a VP/ VC basis. When ATM traffic is transported across an MPLS network, each service category is tunneled through an MPLS LSP with the equivalent DiffServ class. Typically, CBR is mapped to EF and UBR to BE, whereas VBR and UBR+ are mapped to the chosen AFxy class.

Network Resilience High network uptime is critical for a service provider. Building resilience comes at a cost that is highly dependent on the mechanisms used to improve the net work reliability. Therefore, it is vital that the service provider specify the reliability needed. Highpriority services deserve faster protection mechanisms, whereas lower-priority ones can rely on slower alternatives or possibly no protection at all. With the Tellabs 8600 system, connections can be protected in different ways. Obviously, individual individual links can be protected bet ween two network elements. But this same level of protection would apply to all traffic classes using the link. MPLS provides a protection mechanism that can be used use d to enable much finer granularity granularity.. Using these protection mechanisms, individual LSPs can be protected across the network or even along a selected path in the network. For instance, only paths that are carrying certain traffic classes could be protected across the network through allocation of dual paths. This could represent only a fraction of the interface capacity and makes efficient use of the available bandwidth. By contrast, paths with lower-priority traffic classes can be protected such that the recovery time in the event of failure can be a bit longer, whereas Best Effort traffic normally does not need any protection mechanisms and can tolerate some service breaks in the event of a network outage.

The Tellabs 8600 system in wireline transport The Tellabs 8600 system is adaptable to various applications and enables mobile operators to broaden their service portfolio into wireline services. A wide range of interface technologies with service intelligence, plus a superior network management system, enables the service provider to build a single platform that meets both current and emerging business needs. A single upgradable platform and one network management system is much more costeffective than building parallel platforms to satisfy different service needs. The Tellabs 8600 system is highly flexible. It can be used to connect end users to multiple services with very different requirements requireme nts simultaneously. Multiservice delivery is more efficient for the service provider since the same physical network and business management processes can be applied to many services. In addition to wireless applications, the three main wireline service applications that can be implemented with the Tellabs 8600 system are:  E Ethernet thernet  IP

service s services

VPN ser vices

 Broad Broadband band

Internet access

In a typical service provider network, all of these services can be offered to satisfy the needs of different customer segments or for the operator’s internal use. Tellabs believes that it makes the most sense to utilize the same access and core infrastructure for implementing all services.

Ethernet Services With an Ethernet service, the customer manages the end-to-end routing and the service provider simply provides Ethernet connectivity between each customer site. In most cases, this

provides a lower-cost, more flexible and scalable alternative to traditional leased lines. Large corporations, which have their own IT departments in place, often prefer this type of service since they wish to retain control of the routing network net work inside their company. Using MPLS, the Tellabs 8600 system can deliver Ethernet services in two ways:  As

a point-to-point Virtual Private Wire Service (VPWS)

 As

a multipoint-to-multipoint Virtual Private LAN Service (VPLS)

Due to its simplicity and similarity to traditional leased lines, VPWS is currently the predominant service provider offering. VPLS has gained much interest recently but is still in its infancy in terms of technology and network deployments. However, it offers an interesting alternative for current deployments. For example, LAN interconnection services often use a hub-andspoke topology in which the headquarters acts as the hub site. This kind of network can be created easily using VPWS point-to-point Ethernet tunnels. However, However, each VPWS VPW S tunnel is terminated at a separate port on the hub site switch. VPLS can provide a better alternative since the customer only needs one physical interface to the service and still provides any-to-any connectivity between the sites. With VPLS, the service provider network emulates a big Ethernet switch from the end-customer point of view. All the sites look like they are physically on the same LAN, making service cheaper to deliver and easier for the end user to manage.

IP VPN Services An IP VPN service can be considered the next layer of value-added service over and above basic Ethernet connectivity since it adds routing management to the service. But IP VPN services can also provide the platform for more value-added services, which can give access to additional revenue and greater profitability for the service provider. IP VPNs are particularly attractive to customers with limited IT support skills. They are also requested by companies for whom IT is not a core competency and who wish to outsource as many services as possible. The Tellabs 8600 system implements IP VPNs based on RFC 2547bis. This is the IETF standard that describes a mesh service model for LAN interconnection and makes use of the Quality of Service and Traffic Engineering capabilities offered by MPLS. This IP VPN method uses a peering model in which the customer’s edge routers exchange their routing messages with the Provider Edge (PE) routers. MP-BGP is then used within the service provider network to exchange the routes of a particular customer VPN among the PE routers that are attached to that VPN. This is done in a way that ensures that routes from different customer VPNs remain distinct and separate, even if two VPNs have an overlapping address space. The PE routers distribute the routes from the CE routers to the other CE routers in that particular VPN. This IP VPN model scales well in large customer networks and supports different network topologies, from hub-and-spoke to full mesh. Customer routes are propagated in the service provider network with the help of the MP-BGP routing protocol, which automatically provides updates of the correct VPN routes in the respective PE routers. The Tellabs 8600 system introduces a similar hierarchical model to that specified for VPLS services into the IP VPN application. This is discussed later in this document, along with a comparison with the standard IP VPN models. 15


Broadband Internet Access

Network and service deployments

Broadband Internet Internet access refers to the volume deployment of broadband services to residential, SOHO and SME customers.

With the Tellabs 8600 system, it is possible to create a service oriented network. The Tellabs 8600 system can function simultaneously as a reliable transport network for point-to-point services and a service platform for IP VPNs. In addition, it can efficiently aggregate and transport traffic from DSLAMs and MTUs that generate a massive amount of Internet traffic.

Today, the majority of broadband services are based on digital subscriber line (DSL) services, which use existing telephone-grade copper pairs. Most DSL services are still based on ATM technology, which prolongs the need for ATM in the access network. However, ATM is not seen as a long-term solution and is gradually being replaced with Ethernet and MPLS solutions at the head-end DSL Access Multiplexer (DSLAM). Broadband Internet Internet access can also be offered to Multi Tenant Tenant Unit subscribers. In this case, subscriber traffic is aggregated in the basement of the building with a low-cost Ethernet switch and transported to the service provider network, usually over a fiber connection. This model is becoming b ecoming increasingly common, especially in urban areas and new buildings. Wireless hotspots are also becoming popular in public areas. This too is driving the demand for Ethernet-based transport and aggregation s olutions. The flexibility of the Tellabs 8600 system means that the same platform is suited as well to aggregating traffic from DSLAMs and MTUs to the Internet service provider as it is to delivering enterprise services. In the long term, these services will evolve from their current “Best Effort” requirements to needing true QoS to support IP multimedia and voice services. Because of its carrier-class capabilities, the Tellabs 8600 system is an ideal solution for Internet access deployments, which will have increasingly strict Quality of Service requirements. End-to-end service provisioning is extremely easy and efficient with the provisioning tools provided by the Tellabs 8000 manager. In addition, last-mile connectivity for MTU and wireless hotspot applications can be based on very cost-efficient Ethernet access.

Delivery of Value-added Services The continuing price erosion in basic connectivity services is driving service providers to look for ways to provide higher-value services to their customers. Tellabs Tellabs recognizes that this is one of the key challenges for service providers today. Value-added services involve more than simply offering a flexible SLA. It is the additional services on top of the basic connectivity that can provide profitable revenue sources and help service providers to stay competitive. By the service provider taking on more of the customer’s IT-related needs, a business partnership is created between customer and service provider. This partnership can strengthen the relationshi relationship p over and above a pure bandwidth supply arrangement. If you are supplying only bandwidth, a competitor can always offer it more cheaply. For example, IP VPN services provide an ideal opportunity to add value. Optional services can be added, such as managed firewalls, storage backup services, virus protection, traffic encryption, Web hosting and application management. These services can even be offered by a specialized third-party service provider who leases capacity from the network service provider. The key to offering differentiated services is the ability to treat traffic streams in different ways throughout the delivery network. This is something that the Tellabs 8600 system can do both effectively and efficiently.

16

The Tellabs 8600 system is purpose-built for the service provider environment with a complete set of carrier-class features. It is designed to be cost efficient to deploy for even a small number of services and to be able to grow with the service provider’s business.

Point-to-Point Services (VPWS) In a VPWS, end-user traffic is tunneled through the packet-switched network along Pseudo Wires. An Ethernet PW emulates a single Ethernet link between t wo end-points. Typically, Typically, the underlying network is based on IP/MPLS technology. The most common of the tunneling methods is PWE3, also sometimes referred to as the “Martini draft” implementation.

Figure 13. Ethernet PWE3 tunnel using the Tellabs 8600 system

As shown in Figure 13, the Tellabs 8600 system implementation of VPWS implements the PWE3 draft. The encapsulation of different frames or cells into MPLS labels emulates a leased-line type of connection. The transported traffic can be Ethernet, ATM, FR or TDM. PWs are constructed by establishing a pair of unidirectional MPLS virtual connection LSPs between the PE end-points. One of these tunnels is used for incoming and the other for outgoing traffic. These LSPs are identified with MPLS labels, either assigned statically or provided dynamically using the Label Distribution Protocol (LDP). Ethernet traffic can be mapped to the PW tunnel on the basis of its ingress port or by using its VLAN ID information. An LSP carrying multiple PWs is built across the MPLS network that connects the PE routers. This tunnel can use either LDP or RSVPTE signaling. With RSVP TE, the PWE3 tunnel can have bandwidth guarantees and traffic class characteristics assigned in the same way as with an IP VPN. The inner label identifies the physical or logical interface at the ends of the tunnel connection. This can be the Ethernet port or VLAN ID, the ATM VC or the FR DLCI, depending on the original traffic type. The encapsulated traffic is not examined or inspected by the intermediate routers along the connection. And since all of the information shared along the traffic path is at the MPLS layer, the security of the encapsulated traffic is maintained. In addition to offering similar bandwidth guarantees to those of IP VPNs, VPWS can take advantage of the same MPLS traffic protection mechanisms found in an IP VPN. The Tellabs 8000 manager makes the provisioning of single Ethernet tunnels very straightforward. Even a large mesh of tunnels can be provisioned simultaneously using the same easy-to-use tools.

IP VPN Services In the traditional IP VPN service deployment model, service implementation is the responsibility of the PE router at the edge of the IP/MPLS core. All of the intelligence needed for the IP VPN service resides in that router. The CE router is normally connected to the PE via a point-to-point connection, regardless of the network technology. The CE router is usually an ordinary router, owned by either the service provider or the end customer.

Figure 14. Traditional IP VPN deployment model

Figure 14 shows the basic working principles of an IP VPN based on RFC 2547bis. PE routers are located at the edge of the service provider IP/MPLS core, and traffic from the CE routers is backhauled to the PE router using any of the available access networks. The PE router dynamically peers with the CE router using BGP,, OSPF or RIP routing protocols. Alternatively, the service BGP provider can define a static route in between the CE and PE. The PE router separates the different customer VPNs into logical VPN Routing and Forwarding (VRF) tables. These can be seen by only the corresponding customer part of the VPN. Customer VPN addresses are propagated over the core network using the MPiBGP routing protocol. This inserts the VPN routes in the correct VRF table corresponding to the VPN on the PE routers. Usually in a large or growing network, Route Reflector units are used for BGP communication and scalability. Instead of having a full mesh of BGP communication between all of the PE routers, each PE establishes a session with a Route Reflector, which distributes the routes to the relevant PEs. Normally, the Route Reflector is duplicated and an additional session is established from a PE to the secondary Route Reflector. In this deployment model, a single PE router is typically responsible for offering services to a large number of end customers. The processing power and reliability of the router are therefore critical to the operation of the network. Enlarging the network often requires adding a new PE router: a significant financial and operational investment that needs to be cost-justified on the basis of potential customers and traffic. When some level of redundancy is required, the cost can become even more significant. In summary, the limitations of the current deployment model are that:  Scalability

is limited by the cost and complexity of introducing a new PE router.

 The

cost per bit in the access network is relatively high since the legacy network technologies are not optimized for transporting bursty data traffic.

 Metro

Ethernet deployments lack the required QoS capabilities and hence require heavy over-pro over-provisioning. visioning.

 End-to-end

service management and monitoring is a challenge across the disparate platforms and technologies currently deployed, which often leads to SLAs covering only the PE–PE part of the service.

 Inefficiencies

arise from operation of multiple network net work technologies, such as TDM, ATM, FR or Ethernet in the access network and IP/MPLS in the core.

 Difficulties

o ccur in mapping IP and Ethernet ser vices to ATM or occur FR service models in the access network.

To address address these limitations, Tellabs Tellabs has created a new distributed architecture to support IP VPN services, which makes scaling a network easier and faster. The architecture architecture takes the hierarchical model introduced by the IETF for VPLS services and extends it to the IP VPN, a natural step since networks often will be used to deliver both types of service. The Tellabs 8600 system can be used to deliver the standard IP VPN model as well as this distributed version. In practice, the two models are likely to coexist in the same network: larger areas will be implemented with the new distributed model and smaller areas, with limited growth expectations, using the existing flat model. In the distributed model, the same procedure and protocols that were used in the core between the PE routers are applied in the access domain. In routing terms, distributing the network improves the management, scalability and stability of the network. Traffic Engineering and protection mechanisms can be implemented optimally within the regional networks, regardless of the core network configurations and set-up.

Figure 15. Distributed IP VPN enabled with the Tellabs 8600 system

The distributed IP VPN model is shown in Figure 15. PE routers are divided into U-PE (user-facing (user-facing PE) and N-PE N -PE (network-facing PE) as is done in the hierarchical VPLS specification. The U-PE router has a direct connection at IP level with the CE. The N-PE is at the edge of the core and communicates with the other N-PE routers across the core as in the existing flat model. MP-eBGP is used for VPN route distribution in between the U-PE and N-PE as in the standard model. Where there are Tellabs 8600 system platforms or other IP/ MPLS routers in the network between the U-PE and N-PE, they act as simple MPLS Label Switch Routers (LSRs) just as P routers do in the core. They are indicated as “P-a” (P in access network) in the diagram. P and P-a routers do not need to understand anything about the VPNs since they only transport traffic on the basis of the outer labels. It should be noted that, in a typical network, one element has several roles. For example, for one service the router can be a U-PE and for another a P-a router. The limitations of the traditional IP VPN model can be addressed with the distributed model. Full-mesh connectivity is required only between the N-PE elements in the network. Also, the addition of a new U-PE to the network is more straightforward: it only needs connectivity within the access or regional network. All of the customer VPN routes are communicated in a consolidated manner in the regional network between U-PE and N-PE routers using a single MP-eBGP session. Where customer sites are in the same region, traffic can be locally routed without loading the core network. Thanks to the element architecture, network growth can 17


be achieved with incremental investments making it more economically attractive. Provisioning services with versatile resilience mechanisms over the access network or at the edge of the core adds marginal cost in the distributed model. And MPLS protections, dual homing and Traffic Traffic Engineering can be used in the regional network since it is a traffic engineered domain on its own. The basic principle in the distributed model is to move the intelligence away from the core and closer to the customer demarcation point. The new access domain utilizes the same procedures that are traditionally used only in the core. In practice, this has the added benefit of removing the single point of failure at the edge of the core network. This is all made possible through the cost-efficient architecture of the Tellabs 8600 system. The management components of the Tellabs 8600 system can automatically set up the required LSPs and parameter configurations for the network elements along the VPN route. A CE router or switch in the end user’s LAN is connected to a Tellabs 8600 system U-PE device either on customer premises or at the Local Exchange (LE) site, typically with an Ethernet interface. The service provider connects all of the sites, which are part of a specific VPN, according to end user requirements. The service provider also sets all of the QoS parameters for each VPN according to the end-user requirements. With the Tellabs 8600 system, different traffic types can be classified at the customer demarcation point before entering the operator’s network. Another option is to do this at the first LE site using an Ethernet aggregation switch such as the Tellabs ® 8606 Ethernet Aggregator. Alternatively, this function can be performed by way of a CE router located on customer premises. This might be the preferred solution in cases where high bandwidth fiber is used for the local loop or where true end-to-end management is not an issue. All of the customer traffic is transported over MPLS LSPs in the regional and core networks. This results in a single LSP in the core network and separate LSPs in the regional networks on either side of the core.

The Tellabs 8600 system in combination with its accompanying access nodes supports many types of customer network access technologies, including Ethernet, TDM, DSL and wireless access. Customer networks can be distinguished from each other by a combination of port, channel, circuit, VLAN ID and MPLS label at the Tellabs 8600 system interface used to connect to the chosen access network. Customer traffic entering the network is directed to the appropriate service on the basis of service-specific policy settings for the Tellabs 8660 switch. For example, certain VLANs can be forwarded to a VPWS while others are directed to an IP VPN service. A customer’s broadband Internet access traffic can be directed to a dedicated service network using:  IP

routing where the customer Ethernet VLAN or ATM VC is terminated at an IP router. Traffic conditioning based on the customer SLA and IP address relaying from the DHCP server is performed at the first Tellabs 8600 switch.

 PW

tunneling using either the Ethernet/VLAN ID or the ATM VC, from the DSLAM to the BRAS.

Figure 16 shows multiple service provider networks with a centralized BRAS service selection gateway. These services are provided to customers across a regional network. Traffic leaving the network toward the customer is combined from the multiple service networks. It is then queued and shaped according to the service-specific policies at the Tellabs 8660 switch. Packet replication for predetermined groups is performed to support multicast services such as IPTV. Local content servers and caching can also be supported for additional service networks.

In summary, the benefits of the distributed Tellabs 8600 system in delivering IP VPN services can be:  Improved

scalability: a single PE takes care of a larger number of customers and customer routes are communicated more efficiently with one BGP session across the access domain.

 A

cost-efficient entry point for new networks: the system gives the option of starting with limited services and gradually extending to a large service delivery platform.

 Operational

efficiency in service provisioning and upgrade processes: the Tellabs 8000 manager enables fast service creation. It is easy and accurate to use since the operator does not need to configure each element individually or have a deep technical understanding of each network element.

Figure 16. Tellabs 8600 system in broadband service aggregation

 Better

In a next-generation broadband access architectur architecture, e, the primary application of the Tellabs 8600 system is to support multiple services such as data, voice and video on a converged regional and access infrastructure. Quality of Service and bandwidth usage can be controlled in the metro network, and both business and residential services are supported by the same regional and access infrastructure.

scalability of MPLS Traffic Engineering: with fewer PE routers in the distributed solution, there are fewer tunnels to traffic engineer over the core.

Broadband Service Aggregation All of the applications supported by the Tellabs 8600 Managed Edge System can benefit from the platform’s broadband service aggregation capabilities. In provision of Ethernet, IP VPN, Internet access or value-added services, network traffic can be separated and aggregated at the edge of the network according to the specific service needs. 18

Managed migration path from the Tellabs 8100 and Tellabs 6300 systems

when one is building a network connection, gathering service level data or troubleshooting the network.

In today’s telecommunications environment, it takes time to transition the network to new technologies, services and standards. Current production networks must operate in parallel with new developments in order to maintain revenue streams and maximize the profitability of the existing networks.

Integration with third-party OSS systems is also faster and easier, since only one platform instead of two or three needs to be integrated. Maintenance Maintenance of the management platform is also much easier and less costly because there are fewer components to look after.

Tellabs has made a number of enhancements to its exis ting platforms to support a smooth transition to next-generation IP- and Ethernet-based services and networks. In practice, this means that new services can be introduced quickly and easily with small incremental investments while the established business processes are continuously maintained. maintained. This is especially important for the Tellabs 8100 and Tellabs Tellabs 630 0 system solutions, s olutions, which have a significant global installed customer base with extensive network coverage in both wireline and wireless networks. Ethernet interfaces and switching are already available as add-ons for the Tellabs 8100 and Tellabs 6300 system platforms. An Ethernet interface is the most cost-efficient and flexible way to build connectivity today towards IP-capable devices. With Ethernet switching, the TDM platform can be utilized in the most efficient manner and with more flexible connectivity for multisite networks. The service provider can choose the best option from among E1, STM-1,, Fast Ethernet and gigabit Ethernet when connecting to its STM-1 Tellabs 8100 and 6300 system elements. To make the transition to IP/MPLS as seamless as possible for service providers, Tellabs has ensured that even though the Tel Tellabs labs 8600 and Tellabs 8100/6300 system platforms are based on different technologies, they share a common management system, as shown in Figure 17.

Migration in Wireless Networks In wireless transport networks, a combination of the Tellabs 8100 and the Tellabs 6300 systems is usually deployed for GSM networks. When starting to deploy 3G networks, the service provider has two options: to invest in new Tellabs 8600 system elements or to upgrade the existing elements to provide more capacity in the network. In high-density areas, it often makes sense to start immediately with the Tellabs 8600 system, since it is optimized for ATM and IP and scales easily for future needs. In areas where the bandwidth capacities are not expected to grow rapidly, it may be sufficient to upgrade the existing systems. Both of these scenarios, as shown in Figure 18, give the service provider various options for building the required connectivity. Both 2G and 3G traffic can be transported over the Tellabs 8100 8100 and Tellabs 6300 system units toward the BSC and RNC site. The Tellabs 8600 system solution can also transport both 2G and 3G traffic; hence, the operator can choose the most economically feasible network configuration for each area in the wireless transport network.

Figure 18. Combined 2G and 3G network with the Tellabs 8100, Tellabs 6300 and Tellabs 8600 systems

Figure 17. A single management solution for Tellabs networks

The Tellabs 8000 manager provides a single database with integrated network management tools providing the same “look and feel” for service management functions. The service provider can save on CAPEX and OPEX by continuing to use the same servers and operational environment. environment. The same s ame management logic is retained, which helps customers to learn quickly how to use the new tools and to maintain the same business processes. Training for service personnel is therefore kept to a minimum. Each service and connection or group of connections can be provisioned, managed and monitored end to end regardless of the termination or origination platform for the service. This can be a major benefit

Mobile networks are growing continuously, continuously, and setting set ting up the new cell sites and building connectivity quickly for them is a challenge – especially in cases when many technologies are involved. The Tellabs 8000 manager makes it possible to bring the new sites into use quickly or remotely manage other network and connectivity changes, which can be caused by network growth or maintenance. Connections starting from the Tellabs 8100 portion of the system solution and ending at the Tellabs 8600 system component can be provisioned and end to-end tested with the management system. This is something that normally would require totally separate tools and management systems and procedures with no interaction in between. A technologically complex network becomes simple with the intelligent management software.

19


Migration in Wireline Networks Figure 19 shows the role of the different Tellabs platforms in a wireline service provider network. The integrated Tellabs 8100/6300 system platform is generally used as an access platform for lowerspeed IP VPN, Ethernet service or Internet access tails ranging from n x 64 kbps to 10 Mbps, whereas the Tellabs 8600 system platform is optimized for connectivity and services at speeds of 10 Mbps and above.

Access equipment typically has less capacity than aggregation nodes deployed in the regional network. The Tellabs 8620 and the Tellabs 8630 switches are designed primarily for small hub sites. The Tellabs 8660 switch is more suited to deployment in the regional network for aggregating traffic from the RAN network to the RNC site. Compact and cost-efficient, the Te Tellabs llabs ® 8605 Access ® Switch and the planned Tellabs 8607 Access Switch are optimized for cell site access. The network elements are based on the same technology platform, which facilitates interoperabil interoperability. ity. The managed access solution may be complemented with a compact Ethernet switch that can be managed similarly to all of the other Tellabs 8600 system elements.

Tellabs Te llabs 8660 s witch The Tellabs 8660 edge switch is the largest and highest-capacity network element in the Tellabs 8600 system sy stem family. Usually, Usually, this element resides at large hub sites or next to an RNC within a mobile operator network. However, due to its intelligent hardware hardware architecture, the element can also be cost-efficiently deployed for smaller sites. These are typically sites that have high reliability requirements and growth expectations; they can operate with only a fraction of the platform’s maximum capacity, offering excellent growth potential. Figure 19. Tellabs 8100, Tellabs 6300 and Tellabs 8600 system solutions in the wireline network

To support new wireline services with minimal investment, the Tellabs 8100/6300 system platform can be upgraded with Ethernet interfaces and Ethernet switching capabilities. These allow capacity to be utilized in a more efficient and flexible way on the TDM platform. Aggregating traffic through an Ethernet interface is very cost-efficient when compared to using traditional channelized channelized interfaces. Furthermore, services can be classified and prioritized as well as managed end to end. Different customer services are identified with VLAN identifiers, and they, in turn, can be mapped to an IP VPN service in the Tellabs 8600 system domain. The copper access capabilities of the Tellabs 8100 system are comprehensive and very flexible. This includes high-per formance network terminating units, with an up to 12 Mbps line speed on copper,, that extend full management capabilities to the customer copper premises. They can be used for their basic Layer 1 functionality or extended to use higher, Layer 2 or Layer 3, functionality through the addition of bridging and routing options. The latter is particularly useful for a Tellabs 8600 system based service extension. With a consistent platform and unified management processes, it is easy and cost-efficient to offer new services or implement branch office connectivity with either IP VPN or Ethernet services. In areas where bandwidth demand is currently relatively low and there is an existing Tellabs managed access network, this can offer a fast and low-cost option for introducing new data services.

Network elements The Tellabs 8600 system comprises several network elements and an integrated, service-oriented network management system. The network elements can be located either in the access network close to cell sites or within the regional network for traffic aggregation and service provision.

20

Figure 20. The Tellabs® 8660 Edge Switch

The physical dimensions of the Tellabs 8660 switch are: 440 x 600 x 300 mm (W x H x D). It can be installed in a standard 19-inch rack, with up to three Tellabs 8660 switch elements per rack. Figure 20 shows the front view of the Tellabs 8660 switch, with space for 14 modules. Module slot numbers 1 and 14 are reserved for the Integrated Control and DC Power Feed Card (CDC) with one slot for redundancy. redundancy. The remaining slots are available for a maximum of 12 line cards (LCs). Different types of LCs may be freely placed in any slot between 2 and 13 in the switch. Thanks to the distributed switching architecture, no switch card upgrades or additions are needed – only line cards need to be added to meet the service provider’s specific interface and functionality needs. The backplane contains buses for data, battery, synchronization as well as a fan module control. Each LC and CDC is connected to every LC and CDC via the backplane using point-to-

point connections. Switching is performed on the LCs, while the CDC provides the information that the IFCs require for making their forwarding and switching decisions. To increase the flexibility and scalability of the chassis, every line card can be loaded with up to two Interface Modules as different combinations. Several interfaces support multiple protocols, which make usage very flexible and allow having a mixture of protocols even within one interface. As a result, the entry cost of the device is very low compared to traditional, centralized switch-based architectures. The backplane itself is passive and contains no active components. This distributed switching architecture architecture gives the following advantages:  It

simplifies the card/slot placement rules and can radically decrease the entry cost of the Tellabs 8660 switch network element.

 It

eliminates the potential for a single point of failure in the element.

 It

provides more space for Line Cards that can deliver services and revenue.

Each LC contains an Interface Module Concentrator (or IFC, a sort of baseboard for an LC) plus up to two Interface Modules (IFMs) and provides a bidirectional interface capacity of 3.5 Gbps. The total capacity of the node depends on the number of populated interface slots. When the node is fully loaded, the total bidirectional interface capacity is 42 Gbps. For future scalability, the backplane can handle 10-Gbps Interface Modules, such as those for 10-Gbps Ethernet or STM 64/SONET 192c. Due to the hardware-based design, all traffic can be forwarded at wire speed.

LCs can be easily removed and reconnected with the help of hooks placed at the top and bottom of each card. In the lower part of the network element there are cable ducts, forced cooling modules with filters and an air intake gap. Fan trays are also controllable via the backplane. The Tellabs 8660 switch can operate at temperatures between –5° C and 45° C, which is within the typical climate range of a telecoms equipment room.

Tellabs® 8630 Access Switch The Tellabs 8630 switch is a more compact version of the Tellabs 8660 switch and has physical dimensions of 440 x 230 x 286 mm (W x H x D). Its smaller size makes it ideal for medium-sized hub or traffic aggregation sites in the mobile RAN, where the compact physical size saves on valuable rack space. The element is normally installed in a standard 19” rack. All cards are positioned horizontally so that the power and control functions reside in CDC cards in the bottom and top slots for a fully redundant configuration. In between these, four slots are available for LCs. These can be equipped with IFMs as in the Tellabs 8660 switch. The same cards and interfaces can be used in both the Tellabs 8660 and Tellabs 8630 switch products, making management of spares easier. When all four slots are used for interface cards, the element provides a maximum forwarding capacity of 14 Gbps. The functionality and flexibility of the unit are identical to those of the Tellabs 8660 switch system. Figure 22 below shows the front view of the Tellabs 8630 switch.

Figure 22. Tellabs 8630 switch

Tellabs® 8620 Access Switch Figure 21. Interface Module Concentrator

The Tellabs 8660 switch is fully compliant with carrier-class reliability requirements since it has been built specifically for use in telecoms service provider networks. Not only can the common logic be duplicated for resiliency in the element, but traffic protection can also be added at various layers. MPLS protection mechanisms deliver failover times that are equivalent to those in protected SDH networks. All LCs and CDCs are hot-swappable; if an LC fails, it can be replaced without disrupting the traffic on the other cards. The system automatically takes care of copying the previous parameters to the new LC. Embedded software in the CDC can be upgraded with no impact on the traffic flowing through the element.

The Tellabs 8620 switch uses the same technology as the Tellabs 8660 switch. It is designed to be used in a base station or small hub site and can be installed in a standard 19-inch rack. Depending on the location, the element can be equipped with the required IFMs and AC or DC power options; optionally, DC power supply can be duplicated. The Tellabs 8620 switch, like all other Tellabs 8600 system network elements, is managed and owned by the service provider provider.. The Tellabs 8620 switch can deliver both voice and data services for wireline or wireless applications. It can handle all of the traffic classification and prioritization. Since the Tellabs 8620 switch is managed by the service provider, it is possible to monitor the endto-end service or connection quality. This is particularly important for SLA reporting. Being able to mix the different traffic streams and 21


services in the access network allows the service provider to deliver more services using the same network and hence increase profitability. These different traffic streams can include mobile network connections as well as business services on the wireline side. Each traffic stream can be mapped into specific tunnels or service instances by the Tellabs 8620 switch. This allows the traffic to be identified and delivered with full end-to-end security.

that are part of the Tellabs 8000 manager. The elements have an identical appearance, the only difference being the interface offering. Both products are configured with two gigabit Ethernet interfaces and additionally a fixed number of T1/E1 and Fast Ethernet interfaces. The Tellabs 8605 switch, shown in Figure 24, has a combination of 16 T1/E1 and t wo Fast Ethernet ports, whereas the Tellabs 8607 switch has a combination of eight T1/E1 and eight Fast Ethernet interfaces. The power supply is selectable between 24 VDC, 48 VDC and AC. Due to their typical role in the mobile network next to cell sites, the switches are environmentally environmen tally hardened so that they sustain a wider temperature range than normal telecoms equipment.


Figure 23 shows the front view of the Tellabs 8620 switch. The unit is a compact and modular network element with a bidirectional interface capacity of up to 3.5 Gbps. It integrates all of the common logic, such as power, switching and control functions, within the same element. The two IFM slots can be equipped with a variety of IFMs. The Tellabs 8620 switch uses the same range of Interface Modules available available in the Tellabs Tellabs 8660 switch and the Tellabs Tellabs 8630 switch. The interfaces may be customer-facing or for connecting the element to the network. The service capacity of the Tellabs 8620 switch has limits that can be flexibly specified by the service provider. The service provider can easily upgrade and test each service and individual connections by using the management system when needed. This maintains full control of the network capacity and provides the capability to charge accordingly. accordingly. It also allows upgrades to be kept under control. The network interface must be carefully chosen so that it suits both the infrastructure and the capacity requiremen requirements, ts, at installation inst allation time and in the future. Various types of networking technologies are supported by the Interface Modules. Like all other Tellabs 8600 system network elements, the Tellabs 8620 switch offers a diverse range of network protection protection features, such as LSP Fast Reroute, to meet even the toughest availability requirements.


Tellabs® 8606 Ethernet Aggregator The Tellabs 8606 aggregator is a compact Layer 2 switch that can be managed using the Tellabs 8000 manager in the same manner as all of the other Tellabs 8600 system network elements. It is specifically targeted at network applications where traffic from multiple end users must be aggregated to the Local Exchange site using Ethernet links over a fiber connection. As is shown in Figure 25, the main applications are:  MTU  Port

access aggregation applications

ex tension shelf for Tellabs 8600 IP/MPLS routers extension

The switches are simple to configure, and services can be fully set up and managed using the Tellabs 8000 manager manager.. Where multiple customers each with partially filled interfaces need to be connected to the network, these Ethernet aggregators can offer a very low-cost solution.

Tellabs Te llabs 8605 and 8607 s witches The Tellabs 8605 switch as well as the Tellabs 8607 swicth are excellent for cell site access where a number of E1/T1 interfaces and Ethernet are required in a compact and cos t-efficient form. The elements are primarily optimized for 2G and 3G traffic aggregation but could just as well be used as a CPE when the provider offers, e.g., business services. From the Tellabs 8605 or 8607 switches at the cell site, the traffic is switched or backhauled towards the network and eventually typically to a BSC or RNC through TDM, ATM or Ethernet Pseudo Wires. Regardless of the small physical size, there are full MPLS and QoS capabilities and the maximum capacity towards the network is 150 Mbps. As with the other Tellabs 8600 system elements, TDM cross-connections and ATM switching help to improve the bandwidth utilization. Element configuration, as well as connection provisioning and verification via end-to-end testing, are performed easily via tools 22

Figure 25. Tellabs Ethernet aggregation solution

The Tellabs 8606 aggregator (see Figure 26) supports 24 100BaseTX Fast Ethernet ports and 4 1000Base-X gigabit Ethernet ports, of which two can be replaced with optical SFP connectors. The switch is designed to act as a multiplexer where the aggregate bandwidth is well below the maximum bandwidth of the link. This avoids network congestion along with any consequent impact on service quality. At the first Tellabs 8600 system switch element in the network, all traffic can be classified and any required QoS-related procedures initiated.

The BRAIN handles all of the routine functions for data forwarding and QoS procedures and enables the Tellabs 8600 system to operate at wire speed. This intelligent design plays a significant role in delivering network resiliency mechanisms. It helps service providers to build demanding, QoS-aware services with a high degree of flexibility. A unique feature of the BRAIN is the inclusion of a test generator with which the service provider can test the service or connection functionality. Test Test metrics supported include connectivity, delay, delay variance, packet loss and throughput. These testing procedures can be executed with ease using the Tellabs Te llabs 80 00 manager manager..

Control and Power Card The CDC card is responsible for the following basic functionality:  Control  DC

plane

power feed for the element

 Synchronization

Due to its fundamental role, it can be duplicated to reduce the risk of network outages. Figure 26. Tellabs ® 8606 Ethern Ethernet et aggregator

When the switch is used for port extension, the Tellabs 8606 aggregator is collocated with the Tellabs 8600 system platform to support more Fast Ethernet or gigabit Ethernet interfaces.

Element Elem ent architecture This chapter describes some of the important architectural features that are implemented in the Tellabs 8600 system platform elements.

Hardware-based Forwarding Plane The Tellabs 8600 system has been designed to support a wide variety of services, from business connectivity to mobile transmission and even residential service aggregation. Each of these services has very different requirements, necessitating a combination of hardware-based implementation and a distributed architecture architectur e for functions such as forwarding. Without a hardwarebased forwarding plane, it is not possible to satisfy the wide range of requirements applying for different types of connectivity services. For instance, traffic-aware QoS treatment, specific protection systems and guaranteed bandwidth per application c annot be handled efficiently at software level alone; the volume of packet processing needed would overload a central-processor-based environment. To achieve the best be st combination of performance per formance and costefficiency, all router elements that are part of the Tellabs 8600 system solution rely on the same core architecture. architecture. To implement this architecture, Tellabs Tellabs has designed a custom ASIC: the Broadband Routing ASIC for IP Networks (BRAIN). This is the central building block for all of the customer premises equipment and plug-in units. Within each network element, the BRAINs are connected in a full-mesh topology: the BRAIN in each plug-in unit is connected to every other plug-in unit using point-to-point connections through the backplane. The architecture also allows the use of general network processors to provide differentiated packet processing for added flexibility. fl exibility.

Tellabs has taken a long-term approach in the development of the Tellabs software for the Tellabs 8600 system. The platform control plane implements the latest network protocols, and the layered and modular architecture allows for flexible upgrades. All of the MPLScapable elements that are part of the product family are built from a common software base. From the start, the platform has been designed to allow the easy addition of new features and support the portability of these features to new products. The control plane implements the IP stack functionality, the routing protocols and the configuration function for all routing and servicerelated parameters. The actual traffic forwarding is performed by the hardware-based forwarding plane; any traffic that cannot be handled by the hardware is forwarded to the control plane software for processing. The control plane software supports both IPv4 and IPv6. In order to support QoS aware services and connections, it includes a range of routing and signaling protocols plus traffic extensions. For example:  For

IP routing, the system supports static st atic routing, OSPF(-TE), OSPF(-TE), ISIS(-TE) and (MP-)BGP (MP-)BGP..

 For

MPLS signaling, either LDP or RSVP-TE can be used to build LSPs through the network.

 The

basic IP stack functions include IP forwarding, TCP, UDP, ICMP and ARP modules.

The control plane supports hot swapping, which means that any card can be changed without the need to power down the entire unit. The configuration information from all of the Interface Modules is stored on the CDC. If one Interface Module should fail, it can be replaced with a new one, which automatically copies the parameters from the control card. New firmware versions can be downloaded easily to the control and line cards without disturbing the traffic forwarding. The CDC software can be upgraded without disruption to the current traffic and services. The Tellabs 8600 system software supports graceful restart mechanisms for OSPF, OSPF, IS-IS, LDP LDP,, BGP and BGP with MPLS labels. When two redundant CDCs are present, they copy data between each other while operating. Should a CDC fail, graceful restart allows traffic forwarding to continue. The protecting CDC does not maintain a synchronized set of routing t ables; therefore, the routing 23


tables cannot be used immediately in a failure situation. Instead, the CDC obtains the up-to-date routing information from the network. IP routers are susceptible to various types of denial of service (DoS) attacks. The Tellabs 8600 system has multiple methods for protecting against such attacks. For instance, the system can relieve the effects of a possible disturbance by dynamically restricting the traffic traversing an element. Additionally, all of the incoming and outgoing traffic in an element can be filtered using a hardware based access control list (ACL) that is defined by the system administrator. administrator. The system administrator can also restrict the number of routes propagated by the routing protocols learned from the VRF tables.

Currently, the following Interface Modules are available for the Currently, Tellabs 8600 system platform:  Eight Eight-port -port

Ethernet 10/100Base-TX

 Eigh Eight-port t-port

Fast Ethernet 100Base-X

 Two-port

gigabit Ethernet 1000Base-X

 Eight-port

gigabit Ethernet 1000Base-X

 2+6-port

(2 x 1000Base-X + 10/100/1000Base-TX) Ethernet combo module

 Eight-port

STM-1/OC-3 POS P OS

 F Four-port our-port

STM-4/O C-12 POS STM-4/OC-12

 One-port

STM-16 POS

 One-port

STM-16/OCSTM-1 6/OC-48 48 POS

Interface Module Concentrator

 Four-port

STM-1/OC-3 ATM

For the Tellabs 8660 and Tellabs 8630 switches, the Interface Module Concentrator is a universal baseboard for all LCs. As shown in Figure 27, the IFC can be equipped with two Interface Modules to form a line card. The IFC is 28 mm wide and can hold any two IFMs from the range available. The line card can then be placed in any available Interface Module slot in the network element. The Tellabs 8620 switch can also be configured with two IFMs that are plugged directly into the fixed module slots of the network element.

 Eight-port chE1/chT1 chE1/chT1

The advantage of this mechanism is that there is no need to buy different types of line cards for each service. Use of a single line card type reduces the total quantity of spare parts that must be held in inventory and simplifies the service provider’s field operations.

 24-port chE1/chT1 chE1/chT1  One-port

Multiservice

Multiservice

chSTM-1/chOC-3 chSTM1/chOC-3 Multiservice

The platform is open for broadening to new interface types, which are added on the basis of customer requirements. All of the optical Interface Modules can be equipped with standard SFP (Small Form-Factor Pluggable) Pluggable) connectors, responsible for transmitting and receiving the optical signals. This modularity means that interfaces can be upgraded when needed and supports a “pay as you grow” approach. Multiservice interfaces offer further flexibility for the service provider since a single interface can be configured to carry a mixture of protocols.

Quality of Service Management The Tellabs 8600 system uses IP DiffServ mechanisms for QoS management. When non-IP protocols such as ATM service categories are transported over the network, they are mapped to IP DiffServ traffic classes. In this way, end-to-end QoS can be provided in a transparent manner. Certain components are essential for delivering Quality of Service in an IP/MPLS network. The following features must be taken into account and are supported in the Tellabs 8600 system design:  The

network elements must support the Traffic Engineering extensions of the IGP routing protocols, such as OSPF-TE or ISIS-TE. These extensions are used to advertise, for instance, the link bandwidth for each traffic class.

 Support

should be provided for the Constrained Shortest Shor test Path First (CSPF) algorithm, which can select the most feasible paths in a network, utilizing the network-related information from the TE-enabled routing protocols.

Figure 27. Line card with two Fast Ethernet Interface Modules

 Label

distribution and signaling with RSVP-TE is required when specific resources of a network need to be reserved. With DiffServ aware Traffic Engineering, it is possible to reserve capacity on a traffic class basis. This provides QoS for premium services.

 The

connection admission control mechanism can check that resources along the path can be allocated before the reservations are made. If a link does not have the bandwidth available in the requested service class, then the request is rejected. The use of CAC is optional and can be set up on a link and service class

24

basis in each network element. The LSPs created with the LDP protocol are never subjected to CAC.  Support

for Strict Priority and WFQ scheduling enables efficient delivery of real-time, premium data and Best Effort services in a single network. With advanced traffic conditioning features such as policing and shaping, it is possible to define CIR, PIR, CBS and PBS settings for each service. In fact, CAC combined with priority-based queuing forms the key component for hard QoS.

Figure 28. Scenarios for QoS implementations in wireless networks

As shown in Figure 28, a service provider utilizing the Tellabs 8600 system can use either L-LSPs or E-LSPs to carry differentiated service classes through the network. Either LDP or RSVP-TE can be used to signal the LSP through the network. The Tellabs 8600 system supports both L LSP and E-LSP signaling. The Tellabs 8600 system is able to forward traffic on the basis of traffic class to the correct LSPs. When E-LSPs are used, multiple traffic classes can be carried over the same LSP. The service provider may decide to carry certain traffic classes in one E-LSP and others in a different E-LSP, which can mean that not all E-LSPs in the network are equal. For instance, traffic classes for data traffic may be carried in one E-LSP and all delay-sensitive traffic traf fic in another E-LSP. E-LSP. LSPs can be provisioned automatically with the Tellabs 8000 manager.. The Tellabs 8000 manager 800 0 manager supports automatic updating of the LSPs in the service provisioning process. If the LSP for the required traffic type already exists, its bandwidth can be increased if so required. This simplifies the workflow and reduces the number of errors possible in the service provisioning phase. Manual route set-up is allowed also. If resource reservations are not needed or the traffic-engineered paths are not available, the LDP can be used for setting up the path. When explicit resource reservations are required and Traffic Traffic Engineering is enabled, LSPs should be provisioned using RSVP-TE. LSPs over the core network can be provisioned with the Tellabs 8000 manager in a similar fashion. Again RSVP-TE is used when resource reservation is needed.

standards for quality and reliability. Obviously, any service protection has to be justified on a cost/benefit basis, so networks are usually built with a mixture of various protection mechanisms to best match the individual service requirements. The Tellabs 8600 system can meet even the most demanding service requirements through a mixture of element-level resilience and net work-level protections. In the Tellabs 8600 system, all of the internal buses through the backplane are protected. These include power, battery and auxiliary voltage, as well as synchronization buses. The backplane is passive, which means that it is highly reliable. The data links between all of the line cards are also duplicated, providing two serial buses between each card. The slot for the CDC can be protected by equipping the chassis with two CDC cards. When the unit is thus protected, both the DC feed for the element and the synchronization are protected, in addition to the control plane functionality. When one of the two units is in active mode, the other is in passive or standby mode. However,, the st andby unit holds identical information to the active However one. The software continuously controls the state of the units and dynamically makes changes if needed. Any changes made do not affect the data traffic flow. The CDC includes graceful restart mechanisms for protocols such as OSPF, BGP, BGP with MPLS labels and LDP. LDP. These mechanisms are critical for serv ice provider networks that carry services with high availability requirements. Graceful restart helps to minimize the impact of a routing protocol failure on traffic forwarding; for warding; the forwarding plane continues working for a certain time even though there is a problem in the control plane. All cards in the network element are hot-swappable. When a card is changed and replaced with an equivalent one, all of the original parameters are automatically copied to the new card. It is also worth noting that a failure in a single line card in the system does not have an impact on any other traffic in the network element. The control unit maintains up-to-date information on all of the parameters of the line cards, which can be requested when needed. For network protection, the Tellabs 8600 system supports both linkand MPLS level protection to provide very fast recovery times. At the lowest level, the Tellabs 8600 system elements support Multiplex Section Protection (MSP) for SDH interfaces. With MSP MSP,, one line is reserved for protecting an identical line. This method is called MSP 1+1 or APS protection and provides very fast SDH-layer protection with less than 50 ms of switch-over time. MSP 1+1 and APS protection options for SDH interfaces are enhanced with equipment protection. In practice, this means data travel via an interface on a separate line card to the one offering the protected connection. With Ethernet interfaces, link aggregation can be used to provide two basic benefits. Firstly, it can be used to increase the link capacity by combining several physical Ethernet links to form a higher-capacity higher -capacity link bundle. Secondly, Secondly, if one of the links in the Ethernet bundle fails, the traffic can automatically be transported over the remaining links. LSP protection can be implemented in several ways supported by the Tellabs 8600 system:  RSVP-TE-based

Resilience The Tellabs 8600 system platform has been designed from the outset to maximize network reliability and to conform to SDH

 1+1

1:1 LSP protection

LSP protection based on MPLS OAM (ITU-T)

 BFD-based

1:1 LSP (IETF) protection

Which is the best protection mechanism depends on the

25


interoperability requirements as well as on the required network interoperability availability. availabil ity. The RSVP-TE RSVP-TE-based -based protection enables an approximate switch-over time of 400 ms to three seconds for the selected LSP in the network. The other three alternatives are capable of meeting the standard SDH 50-ms switch-over time.

displayed graphically in the Network Editor tool of the Tellabs 8000 manager.. This information can be manager b e used for Tr Traffic affic Engineering purposes for LSPs routed through the Tellabs 8600 system and third-party network elements.

Path protection based on RSVP-T RSVP-TE E is implemented using RSVP Hello messages, providing a 1:1 protection mechanism. mechanism. This means that in normal circumstances the traffic is only sent along the working LSP. If a failure is detected, the traffic is forwarded to the protecting LSP. LSP. The switch-over swi tch-over time in this case is highly dependent on the frequency of RSVP Hello messages sent between the end-points of the protection group.

Route Reflector

MPLS-OAM-based 1+1 LSP protection uses MPLS OAM packets, which are sent at a given frequency over the LSPs. L SPs. Transported Transported traffic is sent to both LSPs simultaneously, and the receiving end selects the best source. The OAM packets inform the remote end about the prevailing path conditions. The OAM p acket sending frequency can be set by the service provider. When a 10-ms frequency is used, a 50-ms switch-over time can be achieved. A lower OAM packet frequency results in a longer switch-over time. The benefit with this model is that the intermediate nodes do not take part in the protection mechanisms. When the working and protecting LSPs are terminated by different line cards in the same element, protection is provided also against the potential failure of one line card. Bidirectional Forwarding Detection (BFD) is a protocol that can Bidirectional detect faults in the bidirectional path between two forwarding engines. It operates independently of media, data protocols and routing protocols. One potential application of BFD is to monitor the availability of an MPLS LSP. As such, BFD is a lightweight protocol that can be used to detect a data plane failure in the forwarding path of an MPLS LSP.

Management Plane All of the elements of the Tellabs 8600 system solution implement a full range of SNMP MIBs and support command-line interface (CLI) network management applications as well as the GUI-based Tellabs® 8000 Network Manager. However, within the Tellabs 8600 system solution, the Broadband Management Protocol (BMP) is used for communications between the Tellabs 8000 manager and the elements. BMP was chosen for its scalability, security and flexibility. The software architecture allows simultaneous use of the different management interfaces: SNMP, CLI and BMP. Several concurrent Telnet or SSH sessions can be made to a single element. Both Ethernet and serial interfaces are available on the CDC for local management access of each network element. The Ethernet interface can also be used to build an external management network where required. The management plane is protected whenever the CDC is duplicated for redundancy, redundancy, which is typically the case.

Online Core Network Monitoring Online Core Network Monitoring enables the partial management of third-party network elements in the same management domain with Tellabs 8600 system elements. It collects information on network topology and capacity reservations by means of the OSPF-TE routing protocol. The topology and bandwidth information is then 26

Route Reflector functionality is used to improve the scalability of the BGP routing protocol within an autonomous system. Instead of all routers in the AS running BGP forming a full mesh of iBGP sessions with each other, each BGP router creates a session to the Router Reflector, which reflects the BGP advertisements received from one of the routers to all others. One of the BGP routers in the AS can function as a Route Reflector. Additionally, the Tellabs Tellabs 800 0 manager system offers the unique possibility of dedicating a standalone Linux-based server to operation as the Route Reflector. This solution can increase the BGP scalability further by enabling flexible upgrades in the processing performance of the Route Reflector without reconfiguration of the traffic-carrying network elements.

Network management system Network and connection management are becoming more and more important in today’s networking world. The ability to manage a complex and large network with limited non-specialized resources is essential for service providers. An efficiently designed management system can help to streamline processes and can shorten service delivery and repair times considerably. The network management system for the Tellabs 8600 system follows the same ideology as the widely used and well-received Tellabs Te llabs 8100 system. The Tellabs 800 0 manager is a single platform that can manage both Tellabs 8600 and Tellabs Tellabs 8100 systems network elements, as well as the Tellabs 6300 system network elements. This service-oriented network management system is one of the most important parts of a Tellabs solution. It has been designed on the basis of extensive operation experience and customer feedback gained with the Tel Tellabs labs 8100 system’s previous network manager software. It allows customers with Tellabs Te llabs 8100 and Tellabs Tellabs 630 0 systems network net work elements to continue to use the same management system, which has been extended with new tools for newer applications. The key benefits of this graphical-user-interface-based system are ease of use, scalability and reliability reliability.. Each service type is managed with its own optimized tool. Links between the tools are designed so that operations staff can handle complex tasks without the need for a deep understanding of the network or the management structure. The system has been purposefully designed for ease of use by hiding the complexities of the network behind a service-driven point-and-click interface. In a traditional management environment, environment, setting up each new service ser vice often requires extensive configuration of every network element involved in the delivery of the service. With the Tellabs 8000 manager,, simple actions are translated into a series of commands, manager which are then sent automatically to all relevant network elements. This enables very fast, lower error operations without the need for in-depth understanding of the underlying technology. As shown in Figure 30, the normal CLI has been replaced by a graphical user interface (GUI). This enables service providers to

Figure 29. Easy-to-learn, simple-to-use management system

focus their valuable expertise on the more challenging operational issues. The tools and applications can be launched by operators at remote workstations. In this case, highly customizable user privileges determine the rights of each operator.

In addition to managing Tellabs elements, the Tellabs Tellabs 80 00 manager also collects information on the operation and topology of the core IP network elements for monitoring purposes. This allows the service provider to view the network structure as a unified entity. With this view, the personnel can understand the network status and identify any possible bottlenecks that could affect service delivery.

Benefits of t he Tellabs Tellabs 8000 manager The provisioning of QoS-guaranteed services and connections is a complicated task; it involves many steps and requires up-to-date knowledge of the network topology and resource allocation situation. This work can be done manually by accessing each associated network element using Telnet/SSH and issuing the appropriate CLI commands, but this approach requires an experienced networking expert. It takes a lot of time and is very prone to configuration errors.

Figure 30. Network topology and element management tools

The system has been designed with scalability in mind. It can support very large networks containing hundreds of thousands of elements. As the network scales, so does the number of management servers and workstations. This provides maximum protection and efficiency. To allow integration with the existing service provider network management systems, the Tellabs 8000 manager uses open and standard interfaces. In some cases, the service provider may prefer to monitor the Tellabs 8600 system elements using an existing SNMP manager. manager. For this reason, all Tellabs Tellabs 860 0 system elements support the standard SNMP MIBs. It is also possible to configure the elements via a CLI if this is required.

By contrast, the Tellabs 8000 manager automates these individual steps and provides an umbrella interface for each process. This approach does not need anywhere near the same level of network expertise as the manual method. The end-to-end management of the network and service life cycle is achieved in less time, at a lower cost and with fewer errors. For the service provider, the advantages of the Tellabs 8000 manager can include:  End-to-end

service and connectivity provisioning results in fast time to revenue. Remote configuration, automated processes and service templates reduce the time for delivering the services and remove the need for site visits.

 Advanced

testing tools for connectivity, QoS and throughput provide that high quality services can be maintained for customers. These provide accurate service- and connection-level data for SLA reporting.

27


 Fast

applications or modules. Network management users can select from these only the ones they need.

A

The cornerstone of the system is the Basic Package, which contains all of the tools for planning and building the network, as well as for element management. The element management tools include all functionality that is required for monitoring or configuring elements and their components in a Tellabs 8600 system network. It provides tools for element configuration and element-level fault and performance management. In addition, tools for user privilege management, customer information management management tools and the online help system are included in the Basic Package. All of the functionality is provided via a graphical user interface.

troubleshooting with proactive and accurate fault identification maximizes network net work availability. single management solution for the Tellabs 8100, 8100, Tellabs 630 0 and Tellabs 8600 system platforms is complemented with external products from e.g. 3rd party mobile vendors. Support for TDM, ATM, FR, IP and Ethernet technologies provides an easy upgrade path from TDM to IP.

 Provisioning

can be handled by fewer operators since no special in-depth technology knowledge is needed. The advanced graphical user interface enables fast operations and is easy to learn. The system includes comprehensive online online help to support the users.

A

significant amount of time is saved s aved over the command-line approach, especially for large networks. The number of errors can be reduced dramatically through the use of automated tasks and service templates.

A

central database is maintained by the system and is updated in real time with modifications made by multiple users using different tools. A consistency check is made between the different network elements. The database notifies the user of any mistakes and so prevents faulty configuration data from being entered.

 Service

definitions need only be specified at the top level. The system automatically handles all of the complex element configuration work. Templates make it easy to learn and use these processes.

 More

effective Traffic Engineering can be achieved through network virtualization. Network elements, links and even services can be simulated in the database without updating of the physical hardware involved. This allows different network planning options to be analyzed. For example, the effect of a new service on network congestion can be modeled without any actual changes to the physical network.

 Fault

and performance data are collected from network elements and are associated with individual services and connections. The overall state of a connection can be checked at a glance. The fault management monitoring covers network elements, links between the elements, network-wide parameters and the network management system itself.

 The

entire Tellabs 8600 system based network – devices, configurations, services – is automatically documented in the Tellabs 8000 manager database. For example, service configuration information is stored in the database when a service or connection is provisioned. This is then kept constantly up to date if any changes are made to the configuration.

 The

distributed architecture facilitates that there is no single point of failure. It also means that consistency is maintained between the physical network and its management.

In summary, the Tellabs 8000 manager is designed to be quick to integrate, scalable and easy to use. It can deliver a reliably running network with simple service provisioning and monitoring. The system is equally well suited for managing just the Tellabs Tellabs 8100, 6300 and 8600 system elements or for integration into the service provider’s wider umbrella management system.

System Components The Tellabs 8000 manager is a modular software system. The system functionality is divided into several, separately licensed

28

In addition to the Basic Package, there are a number of network and service level management applications. With the provisioning and testing applications, the operator can configure, test and monitor all of the services and connections in the network on an end-to-end basis. Each service and connection can be tested before it is put into active use. The fault and performance information can be viewed at the service level, which helps the operator to respond quickly to customer issues. All of these end-toend management tools enable an operator to reduce the number of steps and the risk of errors in the service delivery phase. This can have a direct effect on delivery time and customer satisfaction.

Using the Tellabs 8000 manager The basic design principle of the Tellabs 8000 manager is that all actions can be planned ahead of time. They are then implemented when the hardware is available, and activated when needed. Once deployed, the services can be tested to confirm that they are functioning properly. properly. The activation of an element automatically triggers the monitoring function. The following sections illustrate how the Tellabs 8000 manager assists with the daily operations and service-related management tasks.

Service Provisioning Steps The first step is to choose the service or connection end-points. The connection type can be multipoint-to-multipoint or point-topoint. Point-to-point connectivity is implemented as an MPLS Pseudo Wire. This can carry Ethernet, ATM, TDM or FR traffic. The end-points are located in the interfaces/sub-interfaces of the Tellabs network elements. A sub-interface can be specified by a VLAN tag in an Ethernet interface or by an ATM VP/VC in an ATM interface. To speed up the operations, the operator can also create a number of Pseudo Wires at once. This group operation option applies also for connection end-point changes. Once the connection is defined, the operator then specifies the traffic constraints and traffic rate. The traffic constraints include information about the traffic classification rules, QoS requirements and parameters for traffic shaping and policing. These traffic constraints are then used to determine the DiffServ classification of the traffic, plus the routing and capacity reservations for the LSPs assigned to the service. The next step is to set up the newly created connection or the whole service. The service is first created only in the database, using the Tellabs 8000 manager provisioning logic. The operator can then check the results of this before actually implementing the connection on the network hardware. If the connection operation fails for some reason, a descriptive error message is displayed to

Figure 31. Service provisioning tool window

the operator. The Tellabs Tellabs 800 0 manager maintains a list of which configuration steps for the network elements have, and have not, been completed. With this list, the operation can be redone after fixing of the problem that led to the failure on the previous attempt. Alternatively, one can back out of the process completely, with no incomplete settings left on the network elements. Figure 31 shows the main service provisioning window displayed in connection of a Pseudo Wire using the Tellabs 800 0 manager. To test the service or connection, the Packet Loop Testing tool can be launched directly from the service provisioning window. Service testing is a logical step for ensuring that any newly provisioned or modified services are working as expected. Testing at the service level, including SLA-related parameters, is currently unique to the Tellabs 8600 system. Once a service or connection is no longer needed, it can be deleted from the network and from the Tellabs 8000 manager database in a single operation. All configurations related to the deleted service are removed from the network elements.

Packet Loop Testing End-to-end testing of services and connections is one of the most important features of the Tellabs 8000 manager. A similar testing tool, the Circuit Loop Test, is provided for TDM connections built with Tellabs Tellabs 8100 system elements. The Packet Loop Test is made possible by special test and loopback generators and analyzers that are built into the hardware and software of the Tellabs 8600 system network elements. Figure 32 shows the Packet Loop Test window in which the tested service and results are displayed. The Packet Loop Test tool provides answers to questions such as:  Does

basic end-to-end connectivity exist between all or only some of the chosen end-points of the service?

 What

are the values for packet loss, delay and jitter (delay variation) for the connection being tested?

 Is

the provisioned connection able to perform transfer at full bandwidth?

If necessary, the test can be configured to be performed automatically for a specified length of time at a given interval. For

Figure 32. The Packet Loop Test tool enables testing of selected services or connections

example, it can be set up to run every second day between 11:00:00 and 11:00:20 as an additional part of the performance monitoring of a VPN. Test results can be reported automatically using email. If an automatically performed test indicates a problem, an alarm can be generated.

Service-level Fault Monitoring The Tellabs 8000 manager is designed to provide a coherent fault monitoring structure that correlates equipment-level issues with the specific services service s and individual connections involved. Element-level monitoring information is mapped against each LSP and the connections or services carried. This means that when an LSP goes down, the operator has a complete picture of the equipment fault that has caused it and of the impact it has on services. This leads to much faster fault resolution and allows the operator to react to the most critical service issues first. Service management allows the operator to monitor faults for a group of objects as one service entity. The group can consist of, for instance, connections terminating at a specific element, elements that are part of a special network or have a special role, or trunks leased from another service provider. This feature makes it possible to pay specific attention to certain parts of the network that could require more attention or quicker responses.

Performance Monitoring In the Performance Monitoring component, an ex tensive array of metrics related to the performance and traffic characteristics of the links and LSPs is gathered from the network. The information also includes class-based packet statistics to provide a higher higher-granularity -granularity picture of network performance. The information collected is normalized and stored in the database. The performance management GUI tool includes a basic set of reports that the user can generate. All stored historical data can be viewed with the tool, for analysis of the most utilized links and LSPs in the network. This helps the operator know when a link in the network needs to be upgraded. The graphical tool can be used also to monitor the performance of a link in real time when one is diagnosing problems.

29


With trend lining, the operator may use historical data to predict when some links in the network need to be upgraded. The performance data can also be exported to external systems for further analysis. The Performance Monitoring tool is very important in a packetbased environment, for keeping link utilization at acceptable levels. It is a key function for compliance with end-user Service Level Agreements.

Web Reporter Tellabs Web Reporter offers online information on the network and services through a standard Web browser (as illustrated in Figure 33). The tool is easy to use for service provider personnel who need to read current status information or obtain reports at various levels concerning the Tellabs Tellabs network. There are a number of predefined report formats that show the network information in HTML form. At the client end, no special tools are required, only a Web browser. In the network management network, instead a separate server is needed that gathers the information from the database part of the Tellabs 8000 manager and converts it into the appropriate format when the report request from the client is generated.

software versions that can significantly increase integration and maintenance costs. The Tellabs 8000 manager provides support for the Tellabs 8600, Tellabs 8100 and Tellabs Tellabs 630 0 systems’ net work elements through a single northbound interface. This means that any element version dependencies are completely hidden by the Tellabs 8000 manager. Therefore, each tool that is part of the Tellabs 8000 manager platform always has the latest information available. Tellabs has many years of experience of integration with a number Tellabs of third party OSS systems, including Micromuse Netcool, Cramer, HP TeMIP, NetCracker, EliteCore and Servion. In addition, some of our customers have successfully integrated the platform with other vendors’ solutions, such as Orchestream O rchestream and Concord.

Management Solution Components Figure 34 shows the management system, which consists of a number of servers and workstations connected to the same Local Area Network.

Figure 34. Management network for the Tellabs 8600 system

Figure 33. Tellabs Web Reporter client

Management System Interoperability The Tellabs 8000 manager provides flexible interface options for communication and integration with other vendors’ Operational Support Systems. The architecture supports open and documented northbound interfaces with flexible communication protocols for OSS integration. A standards-based s tandards-based Java client library is available for easy access to the Tellabs 8000 manager from any platform. The architecture also provides an open communication API for use if other integration options are preferred. In practice, customization is usually needed in any OSS integration project. However, the design of the system architecture makes this integration simple and straightforward. The Tellabs Tellabs 8000 80 00 manager provides high-level northbound interfaces that communicate at the service level in order to hide the lower-level complexity of the Tellabs 8600 system elements and the underlying network. This shields the integrator from the individual element management systems and any changes between different 30

The management network is connected to the Tellabs 8600 system network elements through one or more communication servers. If there are also Tellabs 8100 8100 and Tellabs 6300 630 0 systems elements in the same network, they will require communication servers of their own. The number of communication servers required depends on the network size, the number of connections and the required reliability level. Should a server fail, the remaining servers automatically take over the communications on behalf of the failed one. The network management system can scale to very large networks with tens of thousands of network elements. One communication server can typically handle a network with up to 500 elements. All of the information from the network and services is stored on a central database server. Information can be extracted from the database even by external systems. The database server contains all of the necessary data for the network. This can include elements from the Tellabs 8100, the Tellabs Tellabs 6300 63 00 and the Te Tellabs llabs 8600 860 0 systems. The network is configured in the database, and the configuration is downloaded to each element. In this way, configurations can be checked for errors before the configuration information is downloaded to the network element.

The management server runs the processing logic for the Tellabs 8000 manager tools, ensuring the proper ordering of all operational actions. The management servers can be duplicated to increase network management system scalability s calability and improve availability. Operations staff can access the network management system from workstations. Using the workstation, the operator can run all of the licensed tools for network design, element configuration, service provisioning, service testing, fault monitoring and network performance monitoring. The service provider can delegate the operational responsibilities between different groups by flexibly defining the privileges for each person or group concerned. The number of workstations needed depends on the number of concurrent users. Each can be connected to the management LAN directly or over a Wide Area Network. The workstations operate as thin clients with the processing logic hosted centrally. The network management system enables the operator to easily monitor and configure the network elements and their respective services. If communication is lost between the management system and the network, these functions are suspended while the connections in the network are not affected. All settings in the network elements remain unaltered until communication communication with the network management system is restored. The management system for the Tellabs 8600 system is built on the basis of open, standard interfaces. These interfaces are also used for communication within the Tellabs 8000 manager manager.. Open interfaces also enable easier integration at several levels with the third-party management systems that a service provider may be using. If required, all of the Tellabs 8600 system network elements can be monitored with SNMP-capable systems. Standard MIBs are implemented in all devices. Configuration can be carried out via a CLI also, if necessary. Whichever method is used, the system maintains consistency between the database and the Tellabs 8600 system network elements to provide that all information is up to date. The Route Master is a separate network element that provides two important functions – the Online Core Network Monitoring and Route Reflector as described earlier. The Router Master can be configured to perform either both functions or only one of them. As mentioned previously, previously, any Tellabs 8600 system network element can also function as a Route Reflector. If the Route Master is used as a Route Reflector, it should always be duplicated, regardless of network size, due to the criticality of its role. The Route Master server operates on a Linux server platform. Windows 2003 and UNIX are available for the database server. All of the process logic servers and workstations work in a Windows 2003 environment. The computers associated with the Tellabs 8000 manager can also be installed in a Tellabs 8100 8100 system sy stem network management configuration. For a current Tellabs 8100 8100 system sy stem network manager customer,, the introduction of the Tel customer Tellabs labs 860 0 system and related management tools is a simple task. All network data can be integrated into a single database, with one tool used for monitoring the network. All of the new tools can be run according to the same business procedures used with the existing workstations.

Scalability The Tellabs 8000 manager is designed to scale with the network size, number of users and required management capabilities. The Tellabs 8600 system is designed for use in large national and even international networks, which may consist of tens of thousands of network elements, so network scalability is obviously a critical issue. End-to-end manageability for every single connection and service in the network is preserved as the network grows. The Tellabs Tellabs proprietary communication protocol BMP facilitates minimal error network management operations since it provides very efficient communication between the network elements and the management system. The BMP protocol is similar to the proven DXX protocol, used use d for communication with the Tel Tellabs labs 8100 system elements. As the number of people operating the network increases, it is possible to add more workstations to meet the network management needs. The service provider can also install additional servers in the management network to add management capacity or for redundancy. In the latter case, the additional servers can share the management load and make the system even more faulttolerant. The Tellabs 8000 manager software includes several optional applications, thus allowing service providers to choose the right set of tools for their specific management needs. When a network is small or a limited number of services are provided, the service provider can start with the basic management functionality. As the service portfolio is enhanced or the network grows, a wider selection of applications can be deployed.

Security for Network Management The protocol used to communicate with the Tellabs 8600 system network elements can support very large networks and handle a vast amount of information. The management traffic uses the UDP protocol at the transport layer. Since the management traffic is mixed in with the live data traversing the network, it is important to make sure that the management commands can always get through, even when the network is busy. This is facilitated by giving the management traffic a very high priority. It is also essential to make sure that the management traffic and the associated applications are accessible only to authorized personnel who have the appropriate security clearance. The following security features have been implemented in the Te Tellabs llabs 800 0 manager:  The

service provider’s administrator can define the user lists and specify the privileges for each user. Privileges can be assigned for, e.g., a certain application only. For each application it is possible to further limit the information available or the rights to run certain tasks.

 No

server or workstation software may be started without entry of a valid username and password. pas sword.

 Each

user is authenticated upon login to the system. The login and all subsequent management actions performed by the operator are tracked and logged on the database server.

 IPSec

can be used to secure the connections between the network management system and the network elements.

31


Figure 35. Tellabs 8000 manager platform integration

Management of Services across Tellabs Platforms

Conclusions

Tellabs eases the transition from traditional leased-line services to new packet based services with its integrated approach to network and service management tools. Simultaneous support for new platforms and existing deployments helps service providers to minimize their capital and operational expenditure. A service provider that already provides services using a Tellabs 8100/6300 systems platform can deploy a Tellabs 8600 system and immediately offer new services and connection methods. Ethernet and IP VPN services or ATM connectivity can be offered using the same management system infrastructure, processes and personnel.

The Tellabs ® 8600 Managed Edge System can be used to provide access and regional aggregation for next-generation mobile and converged networks. For a mobile operator, the Tellabs Tellabs RAN solution is a very cost-effective and versatile solution alternative to ATM-based RAN networks. When used in a mobile RAN, it allows the service provider to migrate from one access technology to another at its own pace. This allows a gradual transition between release phases in implementing a 2G-to-3G evolution toward an allIP RAN. Its modular architecture, versatile interface support and scalability make a Tellabs 8600 system solution a potentially good long-term investment.

All new and existing tools can be launched using the same Tellabs 8000 manager toolbox. It is very easy to learn to use the new components because they follow the same logic and have a look and feel similar to that of the existing tools. All of the licensed tools are visible and accessible via the toolbox. Information about all Tellabs Tel labs products is stored in a single database, which maintains the consistency of the data. Network Editor is able to show all Tellabs network elements and thus provide a full picture of the network topology. Topology changes and element configuration can b e performed with ease for the whole network. Additionally, customer management and network fault management can be processed for the whole network from a single window. For large networks, it is possible to limit the view to only certain areas or levels of the network at a time. The management system makes it very straightforward to provision and maintain connections and services: each step in the process can be carried out with the same tools. The system automatically takes care of correctly configuring all of the network elements that deliver a part of the service. The network management system automates the process as much as possible and asks for only the relevant parameters from the operator. In the service creation process, user-friendly wizards facilitate building cross-platform connections. This way, the user is guided through the steps that are needed to implement the task. Using the same system, the operator can manage various service types and multiple technologies. With the testing tools, services and connections can be tested automatically when created or on a regular basis. Moreover, faults occurring in any of the services are reported through the same fault management system. 32

The Tellabs 8600 system solution offers a truly convergent platform that can support multiple applications and services across different segments of the customer base. Wireless LAN hotspots and WiMAX are and will remain a part of the network that requires Ethernet connectivity and high bandwidths. The same Tellabs 8600 system platform and elements can be used for efficiently transporting traffic in a mobile RAN, delivering managed IP VPN and Ethernet services to business customers and aggregating Internet access traffic from residential users through various access options. An integrated advanced management solution – the Tellabs 8000 manager allows the service provider to minimize operational expenses as well as improve network change response times. The solution is extremely scalable and offers the same capabilities even if the network grows significantly. The Tellabs 8000 manager supports multiple Tellabs product families and provides customers with seamless management across platforms, independent of the underlying technology.

Summary Summa ry of product features f eatures

Performance

QoS functionality

The table below provides a summary of functionality and products of the Tellabs ® 8600 Managed Edge System. It should be noted that some parts of the system or functionality listed below are not yet generally available, but are part of the planned future development.

 Tel Tellabs labs

86 60 switch: 8660 swi tch: 93.6 Mbps

 DiffServ,

 T Tellabs ellabs

86 30 switch: 31.2 Mbps 8630

 Tellabs

8620 switch: 7.8 Mbps

Applications  Tr Transport ansport

for 2G and 3G Radio Access Acces s Network with ATM IMA, ML-PPP, ATM VP/VC switching and TDM crossconnections at DS0 level

 PWE3

tunnels (Ethernet, ATM, HDLC, FR,

Interface Modules  F Fast ast

Ethernet, gigabit Ethernet, multiservice ch. STM-1/OC-3c, multiservice ch. E1/T1, STM-1/OC-3c POS, STM-4/OC-12c POS, STM-16/OC48c POS, STM-1/OC-3c ATM

 Possibility

of using all Interface Modules in the Tellabs 8660, the Tellabs 8630 and the Tellabs 8620 switches

TDM)  IP

VPN (RFC 2547bis)

 H-VPLS

(Lasserre-Vkompella IETF draft)

 B Broadband roadband

service aggregation

Physical dimensions (W x H x D)  Tellabs

8660 Edge Switch: 440 x 600 x 300 mm

 Tellabs

8630 Access Switch: 440 x 230 x 286 mm

 Tellabs



860 6 Ethernet Ag gregator 8606 gregator:: 440 44 0 x 44.5 x 300 mm

 Tellabs


Switching capacity  Tel Tellabs labs

866 0 switch: maximum of 42 Gbps bidirectional switching capacity (12 IFCs each with 3.5 Gbps switching capacity)

 T Tellabs ellabs

86 30 switch: 8630 s witch: 14-Gbps bidirectional switching capacity 8620 switch: 3.5-Gbps 3.5- Gbps bidirectional switching capacity

IP/MPLS protocols  Static

routing, OSPF-TE, IS-IS-TE, (MP)BGP4 (RFC 1771 and RFC 2858), LDP (RFC 3036), RSVP-TE (RFC 3209)

 PI PIM-SM M-SM

Ipv4 Multicast

Resilience

DiffServ aware MPLS Traffic Engineering (RFC 3270 and RFC 3564)

 CB CBR, R,

VBR, UBR+ and UBR ATM service categories

 ATM

Forum Traffic Management 4.1

 L-LSP

and E-LSP support for MPLS

 Queuing

for up to eight QoS classes per port for DiffServ traffic and 1000 additional queues for user-selectable services (e.g., VLANs) per IFC

 Traffic

classification classif ication based on protocol, source and destination address, source and destination port and Type of Service (ToS) field

 P Policing olicing

with Two Rate Three Color

Marker  Qu Queue-based eue-based

RED, WRED and tail drop for congestion control

 SP/WFQ

scheduling

 Op Optional tional

traffic shaping per queue

Management

 Common

logic 1+1 protection, hotswappable plug-ins (Tellabs (Tellabs 8660 86 60 and Tellabs Tel labs 8630 86 30 products)

 1+1

MSP (APS), 1+1 (MPLS OAM) and 1:1 (RSVP-TE or BFD) LSP protections, Ethernet link protection

 OSPF,

BGP, BGP with MPLS labels and LDP graceful restart mechanisms

 Element,

network and ser vice management with Tellabs 8000 Network Manager

 CLI

as an option for element configuration

 SNMPv2

MIB support in network

elements

Power requirements Synchronization  SE SEC/Stratum-3 C/Stratum-3  External

clock input and output

 Synchrono Synchronous us  Adaptive  IEEE

timing module

Ethernet

synchronization

1588 Precision Time Protocol

 Clock

distribution capability

 48-VDC

power for the Tellabs 8660 and Tellabs Te llabs 8630 863 0 products

 100

… 240 VAC or 48 VDC for the Tellabs Te llabs 8620 switch

 Universal

VAC or 48 VDC for the Tellabs 8606 aggregator

 48

VDC, 24 VDC or 100 … 240 VAC for the Tellabs 8605 switch



8606 86 06 aggregator ag gregator:: wire speed on all interfaces, 12.8-Gbps switching matrix


8605 860 5 switch: 300-Mbps 30 0-Mbps forwarding capacity

Environmental conditions  ETS

300 019019-1-3 1-3 Class 3.2 3. 2 (In use)

 NEBS

GR-63- CORE (In use)

33


Acronyms and initialisms ACL

Access Control List

FMS

Fault Ma Management System

AF

Assured Fo Forwarding Di DiffServ PH PHB

FR

Frame Relay

APS AP S

Automa mattic Pr Protec ecttio ion n Sw Swit itc ching

GE

Gigabit Ethernet

ASIC AS IC

Appl Ap plic icat atio ion n Spe Speci cifi fic c Int Integ egra rate ted d Cir Circu cuit it

GPT

General Problem Type

ATM

Asyynchronous Tr As Transf sfe er Mod Mode e

GUI

Graphical us user in interface

BE

Best Effort

HDL HD LC

Hig Hi ghh-Le Levvel Dat Data a Lin Link Con Contr trol ol

BFD BF D

Bid idir irec ecti tion ona al For Forwa warrding De Dettec ecttio ion n

HTML HT ML

Hype Hy perrTex extt Ma Mark rku up La Lang ngua uage ge

BGP

Border Gateway Protocol

HSDPA HSD PA

High Hig h Spe Speed ed Dpw Dpwnl nlink ink Pac Packe kett Acces Accesss

BMI

Broa oad dband Ma Management In Interf rfa ace

HSUPA HSU PA

High Hig h Spe Speed ed Upl Uplin ink k Pack Packet et Acce Access ss

BMP

Broa oad dba ban nd Ma Man nage gem men entt Pr Prot otoc ocol ol

IBGP

Internal BGP

BRAI BR AIN N

Broa Br oadb dban and d Rout Routin ing g ASIC ASIC for for IP Netwo Network rkss

IFC IF C

Inte In terf rfac ace e Mo Modu dule le Co Conc ncen entr trat ator or,, in inte terf rfac ace e ca card rd

BSC

Base St Station Co Controller

IFM

Interface Module

CAC CA C

Con Co nnection Admiss ssiion Co Con ntrol

IET IE TF

Inte tern rnet et Engi gin nee eeri rin ng Ta Task For orce ce

CBR

Constant Bit Rate

IGP

Interior Ga Gateway Pr Protocol

CBS

Committed Burst Si Size

IMA

Inverse Multiplexing for ATM

CDC

Control and DC Power Ca Carrd

IMS

IP Mu Multimedia Su Subsystem

CDMA CD MA

Code Co de Di Divi visi sion on Mu Mult ltip iple le Ac Acce cess ss

IP

Internet Protocol

CE

Customer Edge

IS-I IS -IS S

Inte In term rmed edia iate te Sys Syste tem m to Int Inter erme med dia iate te Sys Syste tem m

CIR

Committ tte ed Information Rate

ITUIT U-T T

CLE

Customer Located Equipment

Intern Inte rnat atio iona nall Tel Telec ecom ommu muni nica cati tion onss Unio Union n– Telecommunication Telecomm unication Standardization Sector

CLI

Command Line Interface

L AN

Local Area Network

CORBA COR BA

Common Com mon Obje Object ct Requ Request est Broke Brokerr Archit Architectu ecture re

LAN-IC LANIC

Local Loc al Area Area Netwo Network rk Inte Interc rcon onne nect ctio ion n

CoS

Class of Service

LDP

Lab La bel Di Distribution Pr Protocol

CPE

Customer Premises Equipment

LE

Local Exchange

CPU

Cen Ce ntral Processing Unit

LER

Label Edge Router

CSP CS PF

Con Co nst stra raiine ned d Sho Short rtes estt Pat Path Fir First st

L-LSP

Label LSP

CT

Class Type

LSA

Link-State Advert rtiisement

CV

Connection Verification

LSP

Label Switched Path

DHCP DH CP

Dynam Dyn amic ic Ho Host st Con Confi figu gura rati tion on Pr Prot otoco ocoll

LSR

Label Switch Router

DiffSe Dif fServ rv

Diffe Dif fere rent ntia iate ted d Servic Services es

MAM

Maximum Al Allloc oca atio ion n Mod ode el

DMA

Defferred Ma De Maintenance Al Alarm

MEI

Maintenance Ev Event Information

DS

Dif ferentiated Services

MIB

Management Information Base

DSCP DS CP

Diff Di ffer eren enti tiat ated ed Se Serv rvic ices es Co Code de Po Poin intt

MP-B MP -BGP GP BGP with Mult Multipro iprotocol tocol Extension Extensionss

DSLAM

Digita Dig itall Subsc Subscrib riber er Line Line Acc Access ess Multi Multiple plexer xer

MPLS MP LS

Mult Mu ltip ipro roto toco coll La Labe bell Sw Swit itch chin ing g

eBGP

External BGP

MSP MS P

Mult Mu ltiiple lexe xerr Se Sec cti tio on Pr Prot otec ecttio ion n

ECN

Exp Ex pli lic cit Co Cong nges esti tion on No Noti tifi fica cati tion on

MTU

Multi Tenant Unit

EF

Exp Ex pedited Forw rwa arding Diff ffS Serv PHB

NE

Network Element

EGP

Extterior Gateway Protocol Ex

NMS

Netw two ork Management Sys ysttem

E-LSP

EXP-LSP

N-PE NPE

Netw Ne twor ork k-f -fac acin ing g Pro Provi vide derr Edg Edge e rou route terr

ESW

Embedded sof tware

OAM OA M

Oper Op erat atio ion n, Admi Admini nist stra rati tion on and and Mai Maint nten enan ance ce

EV-DO EV-D O

Code Divi Divisio sion n Mult Multip iple le Acces Accesss Evol Evoluti ution on,, Data Data Only Only

OCNM OCN M

Onli On line ne Cor Core e Ne Netwo twork rk Mo Moni nito tori ring ng

EV-DV EV-D V

Code Divi Divisio sion n Mult Multipl iple e Access Access Ev Evolu olutio tion, n, Data and Voice

OSPF OS PF

Open Ope n Sh Short ortest est Pa Path th Fi First rst ro rout utin ing g pr prot otoc ocol ol

P

Provider router

FDI

Forward De Defect In Indication

P-a

Provider router in acces esss netw two ork

FE

Fast Ethernet

PBS

PIR Burst Size

FEC

Forw rwa arding Equiva vallence Cla lass ss

PDU

Protocol data unit

34

PE

Provider Edge

SP

Strict Priorit y

PHB

Per Hop Behavior

SPF

Shortest Path First

PIR

Peak Information Rate

SPT

Special Pr Problem Ty Type

PLT

Packet Loop Test

STM

Syn Sy nch chro ron nou ouss tr tra ans nsmi miss ssio ion n mo mod de

PMA

Prompt Maintenance Alarm

TCP

Transmiss ssiion Co Control Pr Protoco coll

POS

Packet over SONET

TDM

Time Divis isiion Mu Multiplexing

PPP

Point-too-P Point Protocol

TE

Transit Exchange

PSC

PHB Sc Scheduling Class

TE

Traf fic Engineering

PWE3 PW E3

Pseu Ps eudo do Wi Wire re Em Emul ulat atio ion n Edg Edge e to to Edge Edge

TED

Traff ffiic Enginee eerring Da Data taba base se

QoS

Quality of Service

TLV

Type length value

RAN

Radio Access Network

ToS

Type of Service

RED

Random Ea Early De Detection

TTSI TT SI

Tra rail il Ter ermi mina nati tion on So Sour urce ce Id Iden enti tifi fier er

RFC

Req equ ues estt For For Co Com mme men nts (I (IET ETF F doc docu umen ents ts))

UDP

User Datagram Protocol

RIP

Routing In Information Pr Protocol

U-PE UPE

User Us er-f -fac acin ing g Pr Prov ovid ider er Ed Edge ge ro rout uter er

RNC

Radio Ne Netw two ork Co Controller

VBR VB Rrt

Vari Va ria abl ble e Bit Bit Ra Rate te – re real al-t -tim ime e

RR

Route Reflector

VDSL VDS L

Very Ve ry Hi High gh Da Data ta Ra Rate te Dig Digita itall Sub Subsc scri ribe berr Li Line ne

RSVP RS VP

Res esou ourc rce e Res eser erva vati tion on Pr Prot otoc ocol ol

VLAN VL AN

Virt Vi rtua uall Lo Loca call Ar Area ea Net etwo work rk

RT

Route Target

VoIP

Voice ov over In Internet Pr Protocol

RT

Real Time

VPLS

Virt Vi rtu ual Pr Private LAN LAN Se Serv rvic ice e

SDH SD H

Syn Sy nch chro ron nou ouss Dig igit ita al Hie ierrarc rch hy

VPN

Virt rtu ual Private Netw two ork

1 x RTT RTT

Single Sing le carrier carrier Rad Radio io Trans Transmissio mission n Techn Technology ology

VPWS VP WS

Virt Vi rtua uall Pri Priva vatte Wir Wire Serv Servic ice e

SEC

SDH Equipment Clock

VRF

VPN VP N Routing and Forw rwa arding

SFP

Small Formm-F Factor Plugg gga able

WCDM WC DMA A Wi Wide deba band nd CDM CDMA A

SIP

Session Initiation Protocol

WFQ

Weighted Fair Queuing

SLA

Service Level Agreement

WiMAX

Worldw Wor ldwide ide Inte Interop ropera erabil bility ity for for Microw Microwave ave Acc Access ess

SNMP SN MP

Simp Si mple le Ne Netwo twork rk Ma Mana nage geme ment nt Pr Prot otoco ocoll

WRED WR ED

Weig We ight hted ed Ran Rando dom m Earl Earlyy Dete Detect ctio ion n

SONET SON ET

Sync Sy nchr hron onou ouss Opt Optica icall Netw Networ ork k

XML

Exttensible Markup La Ex Lan nguage

35

36


Statements in this document pertaining to (a) future market or technological trends or developments, (b) future Tellabs products or features, (c) cost-savings, profitability or other commercial or technological advantages arising from a product, service or technology, (d) possible network or system designs or configurations, or (e) other future, speculative or forward-looking statements are for discussion purposes only, subject to change and shall not be construed as recommendations, guarantees or warranties (expressed or implied). Results, outcomes or conclusions may differ.

North America

Asia Pacific

Tellabs One Tellabs Center 1415 West Diehl Road Naperville, Napervill e, IL 6056 3 U.S.A. +1 630 798 8800 Fax: +1 +1 630 798 20 00

Tellabs 3 Anson Road #14–01 Springleaf Tower Singapore Singapor e 07990 9 Republic of Singapore +65 6215 6411 Fax: +65 6215 6422

Europe, Middle East & Africa

Latin America & Caribbean

Tellabs Abbey Place 24–28 Easton Street High Wycombe, Bucks HP11 1NT United Kingdom +44 870 238 4700 Fax: +44 870 238 4851

Tellabs 1401 N.W. 136th Avenue Suite 202 Sunrise, FL 33323 U.S.A. +1 954 839 2800 Fax: +1 954 839 2828

The following trademarks and service marks are owned by Tellabs Operations, Inc., or its affiliates in the United States and/or in other countries: TELLABS ®, TELLABS and T symbol ®, and T symbol®. Any other company or product names may be trademarks of their respective companies. © 2006 Tellabs. All rights reserved. 74.1747E Rev. A 11/06

Tellabs 8600 Overview

Recommend Documents