Gateway opens the way to varying crossconnect definitions
Different meanings for "Sonet gateway" have made digital crossconnects challenging to measure
BERT SOTO
ALCATEL NETWORK SYSTEMS INC.
Along with the emergence of synchronous optical network technology came the problematic digital crossconnect term "Sonet gateway." The difficulty is that product companies define the term differently and each company champions its own definition. In fact, different levels of Sonet gateway functionality exist, and only the most capable crossconnect products are able to achieve and perform at the highest level.
The main emphasis in Sonet gateway functionality is asynchronous-to-basic-Sonet-traffic connectivity. As Sonet deployment increases, this type of connectivity is becoming increasingly complicated because of the additional and different synchronous transfer signal-1s that must interconnect with the same asynchronous signals.
Because widespread 100% Sonet networks are scarce, and will be for some time, most network providers have not begun to fully explore the connectivity problems that arise in full-Sonet networks that use different kinds of STS-1s.
Crossconnect products cannot be considered Sonet gateways if they do not allow crossconnections among different STS-1 payloads. Even with Sonet networks still in the minority, incompatible STS-1 usage is already becoming widespread. To build seamless networks, many STS-1 to STS-1 interconnections will also need to be supported.
Clarifying the issue of Sonet gateway functionality calls for a detailed analysis of its different performance levels.
Basic functionality
Basic Sonet gateway functionality enables network providers to interconnect asynchronous payloads to fundamental Sonet interfaces. The prominent asynchronous interfaces of crossconnect products, just as with most network devices, are available in two bit-rate types--digital signal-1 (1.544 megabits per second) and DS-3 (44.736 Mbits/sec).
But even with only two types of asynchronous interfaces, the variety of available DS-1s and DS-3s complicates network interconnection. For example, DS-1s can possess one of two different line codes--AMI (alternate mark inversion) or B8ZS (bipolar format 8 zero substitution). The DS-1s could also have one of two framing patterns--superframe or extended superframe. With these options, network planners can implement four different DS-1s. Moreover, although less frequently used, TR8-coded (Bell Communications Research Technical Recommendation TR-TSY-000-008) and clear-channel, are other available DS-1s.
A similar situation prevails for DS-3s when dealing with asynchronous issues. The differences, however, stem from how the DS-3 is structured instead of specific coding or framing patterns.
The most common DS-3 in communications networks is an M13 multiplexed version. This type of DS-3 comprises seven multiplexed DS-2s (6.312-Mbit/sec), each of which consists of four multiplexed DS-1s.
Another type of DS-3 is the C-bit multiplexed version. It goes through the same multiplexing procedure as the M13 multiplexed DS-3 with one exception. In the DS-2 to DS-3 multiplexing stage, the C-bits of the seven DS-2s are used for proprietary purposes by network providers. Also, as with the DS-1s, a clear-channel DS-3 is available as a third option.
Basic Sonet gateway functionality involves the ability to take any asynchronous payload and interconnect it to a basic Sonet signal. The most basic Sonet signal is the virtual tributary-loaded STS-1 (51.83 Mbits/sec), which comprises 28 multiplexed virtual tributary 1.5 (1.728-Mbit/sec) signals.
One other STS-1 is included for basic Sonet functionality: The clear-channel STS-1, which has its synchronous payload envelope intact, is added to the virtual tributary-loaded STS-1.
Added connectivity
Unfortunately, there are many types of STS-1s, some of which are incompatible. In other words, STS-1s cannot always be interconnected.
For example, in current networks, transitioning from asynchronous to Sonet-based systems frequently incurs the use of two types of STS-1s. In the loop or access portion of many networks, deployment of optical carrier-3 (155.52-Mbit/sec) Sonet equipment is common. But even with the maturation of these networks, the services that network providers are delivering to their customers continue to exist as asynchronous DS-1s. These Sonet access networks often primarily exist to transport DS-1s.
The DS-1s are accepted by a Sonet fiber-optic terminal and mapped into virtual tributary 1.5 signals. The signals are then multiplexed into STS-1s to build virtual tributary-loaded STS-1s.
Adding to the complexity, the interoffice portion of a Sonet network employs a different STS-1. Sonet networks of choice incorporate OC-48 (2.488-gigabit-per-second) equipment. The primary role of these interoffice facility networks has been to replace such legacy transport equipment as asynchronous lightwave and micro wave radio systems.
Traffic rolls off these older systems in the form of DS-3s to the OC-48 networks. Typically, an M13-multiplexed DS-3 is mapped into an STS-1, resulting in a DS-3-loaded STS-1.
The two STS-1s--virtual tributary-loaded and DS-3-loaded--are inherently different in makeup and cannot be interconnected. Their underlying mappings are therefore incompatible.
Not only must a Sonet gateway network element facilitate those crossconnections needed for basic functionality, but asynchronous traffic must have the capability to interconnect to other types of STS-1s; for example, M13- and C-bit mapped types.
Full functionality
One approach to handling these different mappings contends that products should not be produced and deployed into Sonet networks until user demands force the issue. This approach is short-sighted because network providers continually replace network equipment as networks change--even though it is an inefficient and expensive proposition.
Another approach takes the proactive route. Network providers do not have the experience to predict accurately how and how quickly Sonet networks will evolve. Different technologies, customers and strategies emerge continuously and outdate established plans. Therefore, crossconnect products must be produced to accommodate different network architectures. They would offer network providers a safety net while reducing device obsolescence. These products would evolve with Sonet networks as user demands dictate.
Although Sonet promises seamless networks with end-to-end visibility and virtually limitless connectivity, practical experience dictates that slow progress will rule. Consequently, progressive and flexible equipment is needed to accommodate the inevitable changes that arise in today`s volatile networks. u
Bert Soto is technical marketing specialist at Alcatel Network Systems Inc. in Richardson, TX.