Watch Telephony's Security Webcast Find out what Verizon Business is doing with managed network security and how service providers can cash in with this new service. Watch Telephony's Webcast today.

The access part of the metro network, providing the last-mile infrastructure to residential and enterprise customers, typically accounts for 70% of a carrier’s capital and operational costs. As carriers roll out the next generation of services--voice over Internet Protocol (VoIP), video on demand, IP television (IPTV), gaming, etc.--these costs will only increase. The problem is the inherent inefficiencies in the current access infrastructures, typically based on voice-optimized synchronous optical network/synchronous digital hierarchy (Sonet/SDH) or some enterprise-oriented Layer 2/Layer 3 technologies.

In attempting to evolve their access networks to deliver these new services, carriers are being overwhelmed by the complexity of deploying and, more importantly, managing the resulting networks. Thus, paradoxically, the very goals that the carriers are trying to achieve--reduced cost of ownership, ability to quickly offer new services and scalability--are being undermined.

An emerging network architecture with a more comprehensive, field-tested approach to service deployment substantially reduces operational and capital expenditures. This architecture requires no radical change to existing operational models, is easy to implement and, based on initial feedback, is proving to be an optimal choice for deployments of new, high-demand services. In fact, real-world implementations have produced a nearly 50% decrease in costs compared to today’s conventional architectures, while delivering a high degree of scalability, greater speed to market and an overall competitive advantage.

Carriers have two main customer segments--enterprise and residential users--and demand for Ethernet-based services is burgeoning in both.

Enterprise customers seek to extend local area networks (LANs) across the metro area network (MAN) to support a host of services, and they want to use Ethernet as the transport protocol for many reasons. One, more than 90 percent of available network connections are Ethernet ports, and all client/server systems, personal computers and upper-layer protocol stacks are Ethernet-compatible. Two, enterprise Information Technology (IT) departments are familiar with Ethernet protocols. And three, Ethernet services are less expensive than T1/E1 connections, especially at a per-bit cost.

In residential markets, customers seek a host of new services--for example, video on demand, IPTV and “triple-play” services that bundle data, voice and video in one package--which are also usually based on Ethernet infrastructure.

Legacy Sonet/SDH or Layer 2/Layer 3 infrastructures, however, prevent carriers from capitalizing on the market demand for Ethernet services. While a carrier’s network core is a largely stable entity (the traffic of many customers is aggregated, and individual variances do not greatly influence cost), the access network is more dynamic--a function of the number of customers and specific requirements tailored for individual customers.

The operational and capital costs entailed in managing these dynamic demands of access networks are significantly higher, and that’s why the access network often accounts for such a high ratio of a carrier’s costs of offering services. Any change or failure in the existing Sonet/SDH or Layer 2/Layer 3 access infrastructure can prove very expensive, limiting carriers from quickly, cost-effectively evolving their offerings or expanding services into new areas. Change in a Sonet/SDH infrastructure requires truck rolls; failure in Layer 2/Layer 3 networks take a long time to recover (due to Spanning Tree protocol algorithms) and also use extensive resources.

Security is also an issue. Often motivated by new regulatory pressures, business customers especially are demanding heightened security for their sensitive data. Layer 2/Layer 3 access technologies are especially vulnerable to security breaches of customer data in a shared infrastructure. Scaling in Layer 2/Layer 3 networks is also limited due to inherent virtual LAN (VLAN) limitations (4,096).

To fully capitalize on the business opportunities available to them, carriers require relief from all of these obstacles, plus the ability to maintain base revenue streams. What is needed is a new access architecture with the following design goals:

• simplifying the ability to add any service (voice, data or video),
• security over a shared infrastructure,
• ability to scale significantly,
• carrier-class Operations, Administration & Maintenance (OAM) capabilities,
• interoperability with legacy infrastructures (Layer 2/Layer 3 or SONET/SDH) and
• reduced total cost of ownership (TCO).

The key to engineering a significant architectural improvement is finding the right trade-offs between feature functionality and the cost of the underpinning equipment. An evolutionary, new architecture achieves such a balance by essentially decoupling the access network from the core and optimizing it for service delivery. The access focuses on optimizing the transport infrastructure at a very low cost while offering customers the security and management they have come to expect of world-class carriers.

This architecture proposes an access platform based on native Ethernet; this would offer a converged access mechanism to deliver any service (voice, data or video). As noted, transport devices supporting such services invariably have Ethernet interfaces, and natively carrying these services precludes the cost of conversions and additional latency. Aggregation of the transport further optimizes the use of the infrastructure and significantly reduces operational costs. It is critical to note that the emerging architecture is an Ethernet Layer 1 transport platform--not a Layer 2 Ethernet switch--but with some traditional Layer 2 mechanisms for interoperability and essential value-added features.

By employing well-established tagging techniques, customer data can be uniquely identified and separated over the shared infrastructure; this could be easily enhanced to ensure security of customer data. With mechanisms such as stacked tagging, scalability limitations are eliminated simply while users can maintain the flexibility of configuring their own internal networks. Quality of Service (QoS) assurances can be offered by using simple marking and classification techniques at a frame and packet level. Using Ethernet First Mile (EFM) for OAM, an IEEE ratified standard, management of the access network would preclude the need for the expensive Simple Network Management Protocol (SNMP) overlay networks for management.

What the architecture accomplishes is dramatically simplifying the provisioning of services. It does so by eliminating the cumbersome, expensive configurations necessary in traditional Layer 2/Layer 3 and Sonet/SDH networks and optimizing the use of the physical access infrastructure for most, if not all, of the services being requested by customers. By isolating the access from the core, any issues that arise in a very large portion of the network are managed locally and inexpensively.

Architectures based on carrier-class Ethernet transport platforms have yielded tangible savings--in most cases, more than 50% over existing capital and operational expenditures. The key to achieving this has been understanding comprehensively both customer requirements and the essential features needed to meet them. Too often, carriers have force-fitted solutions into their existing infrastructures, resulting in inefficiencies, high costs and poor scalability.

With complexity largely driven out of the access part of the network, the emerging Ethernet-based service architecture greatly diminishes operational costs, enables smooth introduction of new services and offers carriers a new way of looking at what has become an increasingly difficult issue. Its evolutionary nature offers a realistic approach to carriers positioning for competitive advantage.

Abdul Kasim is vice president, Ethernet business development, with ADVA Optical Networking.

Visit ADVA Optical Networking online.