By STEPHEN HARDY
Spurred by complaints from its members that many in the financial community literally don't know what they're talking about when it comes to optical-network capacity, the Telecommunications Industry Association (TIA) has issued a white paper establishing definitions for network capacity terms. Now, the group hopes by next month to complete a more ambitious task: polling carriers and analysts to determine if the United States suffers from a rampant glut of fiber capacity.
The initial white paper, "Optical Network Capacity and Utilization: Clarifying Terms and Definitions," is the result of what Eric Nelson, the TIA's vice president of global network marketing, called "spontaneous combustion" among some of its members regarding Wall Street's generally negative view of telecommunications and widespread assertion that there is little need for capital expenditures to improve fiber infrastructures. In particular, members felt that much of the discussion within financial circles was based on either misinformation or attempts to compare apples to oranges.
"We realized that there were a couple of issues here that we needed to tackle," says Nelson. "First of all, we needed to try to make sure that the terminology that was used among the vendor community, then among the service providers and others, was consistently used. Because that was one of the things that our members were telling us and that we were reading in the press-that some of the terms were being thrown around rather loosely without knowing what those terms meant."
Most of the terms are basic. "Fiber utilization" is defined as the number of miles of lit fiber divided by the total number of miles deployed. "Channel slot utilization" is the number of WDM channels in use divided by the total number of channel slots shipped. "Average" traffic is the total amount of traffic over a defined period, divided by that period of time.
However, the "peak" traffic definition represents an area where the TIA attempted to provide both a definition and some "how a network really works" information. "Peak," according to the TIA, not only comprises the greatest traffic demand, it also includes the resources necessary to meet "additional network requirements, all of which ensure service delivery to the end user." Thus, network overhead is combined with traffic demand to create "peak," a condition the TIA redefines as "demand for working capacity." It is this demand that carriers architect their networks to meet, the white paper states.
Therefore, when determining how much capacity is used on a network (the "supplied active capacity utilization"), observers should consider the ratio of demand for working capacity to the total "supplied active capacity" of the network. "Supplied active capacity" is the sum of the bit rate of all active channels on all lit fiber. For example, a two-fiber link has an OC-192 DWDM system capable of supporting 10 channels. If one fiber is
lit and only five channels are active, the supplied active capacity is 50 Gbits/sec (five channels at 10 Gbits/ sec each). Observers would then compare how closely the demand for working capacity approached 50 Gbits/sec to determine if more capacity is needed on this link.
In addition to definitions, the TIA also lays the foundation for discussion of numbers generated using its terms. It states that carriers have historically operated at or below 70% active capacity utilization. It also describes scenarios in which the owner of the link above might choose to add capacity by installing a whole new link, rather than by using all channels on the existing 10-slot system or lighting the other fiber.
While Nelson declined to name the vendors that contributed to the white paper, OFS Fitel (Norcross, GA) was a major player. The company had made a point at OFC in March of expressing its displeasure with the current descriptions of overcapacity, including inconsistencies in the use of terms. The company then took its case to the TIA, according to Janice Haber, vice president, systems engineering and market development, at OFS. "Certainly, the definitions that we had proposed [at OFC] are the ones that have been adopted, with very minor modifications," she says of the white paper.
In addition to receiving positive reviews among vendors, at least some corners of the U.S. financial community see a value in the project as well. "When we've talked to folks in the industry-carriers, vendors-as well as our clients, the institutional investors, there certainly has been a lot of misunderstanding around how one measures and what is utilization," says Simon Leopold, a senior analyst for the telecommunications arena at Merrill Lynch (New York City). "So the idea of creating a taxonomy for the topic is a very good one. It makes everybody's life a lot easier."
Putting numbers to terms
What won't be easy, however, is completing the second part of the TIA's two-phased program: putting numbers to these definitions to determine whether there is indeed a glut and, if so, how widespread it is. Nelson expects to release this second report in early September. As of the beginning of June, the TIA was still developing a strategy that would ensure carriers that their proprietary data would remain protected.
However, the organization is determined to proceed. "It's very important now for us to reach out to others who are involved in either using the technology or writing about the technology or in other ways having an impact on the market; it's important for us to reach out to them to make this as complete a report as possible," Nelson says.
Haber agrees, particularly if the TIA's findings paint a positive picture for the industry. "I think the TIA will be able to get a very broad, sort of arm's length perspective on this," she explains. "It won't look so self-serving as it might if one of the vendors, like ourselves, were to come out and say that there's no fiber glut, it's not in all the market segments, etc."
Nelson underscored the organization's technology-neutral stance. "If there is a glut, well we'll say there's a glut. If there's no glut, we will, of course, say there's no glut. And what we might end up with, for example, is that there's a glut in certain parts of the market and not in other parts of the market. But we really want to be as comprehensive and as complete and as objective as possible," he asserts.
While conceding that the TIA's stamp on positive numbers would add to their credibility, Leopold isn't sure that they would have much of an impact on the market. "At this point in the economic cycle and based on what my sense is of investor interest, I'm not sure people care that much," he says. "I don't know that it gives you, from an investment perspective, huge short-term value."
However, Leopold admits that positive numbers would have an effect on his own view of the market, which has been consistently more pessimistic than that of companies such as OFS. "It would certainly be an interesting data point. I would certainly present it, because I think everybody from our perspective, from our audience, is looking for any sign of hope," he says. "And it would cause me to try to figure out what's wrong with my sources of data."