Carrier Ethernet unlocks many potential revenue- generating services that telecommunications service providers, otherwise known as carriers, must deploy to be competitive. However, most carriers aren’t ready to convert to a pure Ethernet network due to Ethernet’s lack of native support for link monitoring, fault isolation, and diagnostic testing.
These kinds of attributes happen to be native to the Plesiochronous Digital Hierarchy (PDH) and synchronous SONET/SDH networks. Over decades, carriers have come to trust PDH and SONET/SDH networks as dependable platforms.
Achieving the transparent and efficient transport of native Ethernet frames from network edge to network edge is a challenge. In the past, overcoming these hurdles has been a rather costly endeavor.
Near the end of the 1990s, many carriers fork-lifted some portion of their networks and replaced them with what was then called “next-generation” SONET/SDH (NGS) equipment. The strength of this equipment was the efficient transport of Ethernet and TDM services when the infrastructure approached 100% utilisation. Its weakness was the lack of interoperability with legacy systems. Today, however, using new protocols that allow the reuse of legacy equipment minimises the overall cost of delivering new Carrier Ethernet services.
Before you can get a handle on the advantages of the new methodology, it’s important to understand a few details of NGS. When transporting Ethernet, NGS solutions place Generic Framing Protocol (GFP) encapsulated Ethernet frames directly into variable-bandwidth concatenated SONET/SDH virtual containers. These solutions primarily used the methods defined by ITU-T G.707.
This transport scheme promised to provide optimal bandwidth usage in a SONET/SDH link when running near full capacity by providing very fine bandwidth granularity for each service on a NGS network. Many carriers regarded this class of equipment as the ideal technological solution.
However, when terminating or handing off a service, these concatenated (linked) virtual containers must be resolved into a physical interface, such as OC-3, STM-1, T1, E1, or DS3. The reason NGS systems don’t interoperate well with legacy systems is the fact that the concatenated virtual containers originating at an NGS node can’t be resolved to a standardised physical interface by a legacy SONET/SDH system.
Because legacy SONET/SDH systems are unable to perform this task, NGS equipment is required at these nodes. In addition, when a legacy network is used to transport a service that originates at an NGS node, typically an entire legacy SONET/SDH container is allocated to the path, eliminating the fibre bandwidth efficiency gained via NGS. In short, NGS systems ignored interoperability with the established transport methods, in favour of bandwidth utilisation promises that were rarely achieved.
The new approach for efficiently transporting Ethernet over SONET/SDH leverages, rather than deviates from, traditional transport methods. To grasp the importance of this approach, we must start with some fundamentals of legacy SONET/SDH systems.
All telecommunications equipment depends on protocol processing in silicon and software to perform the bulk of its duties. The basic protocol stack of a legacy SONET/SDH add-drop multiplexer (ADM) is shown in Stack A of Figure 1. This protocol stack has been used for many years to carry the PDH time-domain-multiplexed (TDM) services, such as leased T1, E1, and DS3 lines.
These PDH services—T1, E1, and DS3—are well understood, globally deployed, and trusted. Therefore, it’s understandable that the International Telecommunications Union (ITU) would adopt these PDH technologies as the transport layer for new Ethernet services.
Recently, the ITU has developed new recommendations for Ethernet transport over single and multiple PDH links. The applicable standards are ITU-T G.7041, G.7042, and G.7043. Collectively, these recommendations are the fundamental building blocks of Ethernet-over-PDH (EoPDH) technology. The protocol stack used in EoPDH equipment is labeled and shown in the top portion of Stack B in Figure 1.
EoPDH is a collection of technologies and new standards that allow carriers to use extensive existing telecommunications copper infrastructure to provide new Ethernet-centric services. EoPDH standards pave the pathway for interoperability and the gradual migration of carriers to pure Ethernet networks. The standardised technologies used in EoPDH include frame encapsulation, mapping, link aggregation, link capacity adjustment, and management messaging.
Common practices in EoPDH equipment also include the tagging of traffic for separation into virtual networks, prioritisation of user traffic, and a broad range of higher layer applications. Although EoPDH was created for point-to-point delivery of Ethernet over physical PDH tributaries, when combined with legacy SONET/SDH, EoPDH becomes an important element and cost-effective tool for Ethernet service delivery.
A new class of SONET/SDH equipment maps Ethernet frames into virtually concatenated PDH tributaries using the EoPDH standards, and then uses traditional mapping techniques to transport the PDH connections over the existing SONET/SDH network. The protocol stack of this equipment is shown in Stack B of Figure 1.
Continued on Page 2.
Because this stack combines EoPDH and PDH-over- SONET/SDH, the technology is called Ethernet over PDH over SONET/SDH, or EoPoS. The division between the legacy stack and the EoPDH stack at a protocol layer compatible with standardised PDH technology allows for an optional physical interface. This makes it possible for the stack processing to be spread across multiple pieces of equipment.
One advantage of EoPoS technology is that it enables a mixed environment of legacy and new equipment. The real strength of EoPoS is that it leverages the existing infrastructure of systems and knowledge for transporting PDH tributaries over SONET/SDH networks. Unlike the NGS approach, which attempted to optimise bandwidth at all costs, EoPoS minimises costs while still making efficient use of bandwidth. To understand these advantages, let’s look at an example application.
In most metropolitan networks, services are delivered over interconnected SONET/SDH rings. One such network ring is illustrated in Figure 2. Although the legacy ADM is diagrammed as a single node in the figure (node C), it actually represents the bulk of telecom equipment deployed in the field. To place appropriate emphasis on the weight of this factor, the installed base of legacy SONET/SDH equipment is literally worth hundreds of billions of dollars.
It’s very important to note that a large portion of this equipment is fully depreciated, and only incurs ongoing operating expense. For a new piece of equipment to decrease total operating costs, the asset’s depreciation expense plus the maintenance expense together must be less than the operating expense of the old, fully depreciated equipment. This single factor makes a strong cost argument for maintaining the operation of legacy SONET/ SDH equipment.
Node A in Figure 2 represents a new piece of equipment using EoPoS technology. In keeping with the principal of interoperability, this equipment typically supports the traditional Ethernetover- SONET/SDH (EoS) and NGS protocols.
Consequently, Ethernet traffic can flow from the new EoPoS node to the NGS system at node B, and from the EoPoS node to the legacy node. As discussed earlier, the legacy node’s protocol stack doesn’t include the NGS protocol.
Because there’s no physical PDH interface in the NGS protocol, the legacy node can’t terminate an Ethernet flow sourced from the NGS node. The legacy ADM at node C is able to transport and hand-off the EoPoS flow from Node A. The legacy ADM processes the bottom portion of Stack B in that’s shown in Figure 1, and provides physical PDH connection to a low-cost piece of equipment. A CPE that supports EoPDH will process the top portion of Stack B in Figure 1, and therefore fully terminates the EoPoS flow.
When an existing customer converts from a legacy TDM service to an Ethernet service, the incremental cost at the legacy node is only a low-cost piece of equipment compliant with EoPDH standards, not an expensive NGS SONET/SDH box. This natural division of protocol processing at the PDH layer is also useful in applications where leased PDH lines are required to reach a customer site where the EoPDH equipment resides.
In addition, when the SONET/SDH network between nodes A and C consists of a complex web of interconnected legacy equipment, the legacy equipment can manage the component EoPoS flows as if they were simple PDH tributaries. While an ADM is used for this example, Carrier Ethernet equipment benefiting from EoPoS technology includes a broad range of equipment types, such as MSPPs, demarcation units, ROADMs, media gateways, IP DSLAMs, and microwave radios.
SONET/SDH equipment enabled with EoPoS technology delivers many of the benefits that were promised by NGS equipment, while optimising deployment expense. By using a standardised virtual concatenation method, the bandwidth consumed by a Carrier Ethernet service can be dynamically adjusted in increments as small as 1.5Mbps.
The ITU-T G.7042 VCAT/LCAS protocol provides dynamic allocation and the flexibility to effectively use all of the SONET/SDH bandwidth. Carrier Ethernet service subscribers can be allocated the bandwidth they require, with little wasted bandwidth. By making intelligent use of the EoPDH protocols in conjunction with SONET/SDH equipment, costs can be minimised while transitioning a network to support new Carrier Ethernet services.
ARTHUR HARVEY is business manager for Ethernet Products at Maxim Integrated Products. He holds a BSEE from Louisiana State University, Baton Rouge, USA.