BY MEGHAN FULLER
While many segments of the telecommunications industry have suffered as a result of the current economic downturn, the optical MAN has continued to see relatively strong growth. The long-haul network has undergone extensive capacity and performance upgrades that have pushed the bottleneck into the metro, where existing SONET infrastructure cannot handle the voluminous data traffic. The need to alleviate this bottleneck has created a wealth of opportunity for vendors. According to a recent report from Pioneer Consulting LLC (Boston), the forecast for the worldwide metro optical-network market is expected to jump from $2.9 billion this year to $13 billion by 2005.
Despite the anticipated strength of the market, carriers and service providers should continue to exercise caution when purchasing transmission systems. According to a new report from Cahner's In-Stat Group's (Scottsdale, AZ) advanced carrier business services, network construction companies are reporting that the cost of building out metro markets may be as much as $50 to $350 per foot. Given this expense, metro carriers and service providers cannot afford to bank on a technology that may not pan out.
Douglas McEuen, senior market analyst of optical networks at Pioneer and principal author of a new report called "Metro Optical Networks: The Emerging Technology Battleground," contends that three technologies are currently "locked in an intense battle to dominate the metro market": next-generation SONET, optical Ethernet, and metro DWDM.
While the long-haul network has undergone tremendous improvements in capacity and performance, "the metro hasn't kept up with the Internet revolution," admits Dan Poranski, vice president of marketing at Tropic Networks (Ottawa, Canada and Andover, MA). "It's still primarily built with TDM types of technology-particularly SONET rings. SONET is great for supporting TDM and voice traffic, but it's structured for very consistent, predictable types of traffic, and that's not what IP and data packets are."
Poranski adds that it's difficult to determine from where the next big wave of traffic will come. As a result, metro carriers have been forced to overbuild their networks ahead of demand, which is a costly proposition.
Startups and traditional SONET vendors alike have begun developing next-generation SONET equipment designed to overcome the limitations of legacy SONET. For example, "vendors are leveraging developments in ASIC technology to expand the capabilities of SONET transport multiplexers to include grooming and crossconnect functionality and statistical multiplexing for efficient data service transport," writes McEuen in his report.
There are clear advantages to staying with SONET-based solutions. Next-generation SONET is derived from legacy SONET, which is well understood and widely deployed. It is also optimized to meet carriers' various quality-of-service (QoS) requirements.
Its detractors, however, claim that these benefits offer only a short-term fix. "As with a lot of legacy products, the problems associated with the protocol are sort of being papered over by the next-generation devices," asserts Agnes Imregh, vice president of marketing at LuxN (Sunnyvale, CA). "The theory is that because there are all these tools in place and because there is all this expertise in place, it pays to stay with the older protocol even though it doesn't have much to recommend it besides that." She argues that while next-generation SONET boxes are cost-saving, they are not enabling.
Yves Hupe, vice president of marketing at VIPswitch (Quebec), agrees, arguing that any SONET-based solution, whether it is DWDM-over-SONET, Ethernet-over-SONET, or another combination, is merely a tactical, short-term solution, rather than a forward-thinking, strategic solution.
"The main point is that the underlying infrastructure remains SONET-based," he contends. "It does not address the long-term issue, which is that most voice traffic will become purely packet-based and that TDM traffic will become marginal."
Interestingly, the state of the telecommunications industry may have created favorable conditions for the deployment of next-generation SONET. "In the last three or four months we have started to see that everyone is kind of backing off from IP, at least for the time being. Now they are saying, 'We'll use our embedded SONET-based networks or TDM-based networks and our ATM-based networks that are already in place,'" admits Dana Hartgraves, vice president of marketing at Metro-Optix (Norfolk, MA).
While McEuen asserts that there is too much SONET in the ground for carriers to abandon it completely, he does admits that "some of the technology advances they've made with metro Ethernet have brought it up to a carrier-class solution to rival SONET." IDC Research (Framingham, MA) predicts that metro Ethernet service revenues in the United States will experience a 36.7% compound annual growth rate, increasing from $155.2 million in 2001 to $740.8 million by 2006.
Like legacy SONET, Ethernet is also tried and true. Optical Ethernet is currently the least expensive technology available for the metro space, and that makes it very attractive to metro carriers and service providers. "We are starting to see even some of the RBOCs look into metro Ethernet," says McEuen. "It's a good solution, and it's viable."
Optical Ethernet does not offer the restoration and QoS guarantees of SONET, though many of its proponents argue that it supplies enough bandwidth to the end user to make QoS unnecessary.
The proponents of optical Ethernet have shown their solidarity in the form of the Metro Ethernet Forum, which boasts more than 50 industry players, including equipment vendors, chip vendors, and service providers. "We believe that there is a strong solution here, and we're going to get together and make this happen," asserts David Yates, vice president of marketing at Atrica (San Jose).
While the metropolitan DWDM market has really only emerged in the last two to three years, it is fast becoming a viable option. As detailed in "Ethernet MAN Services, Filling the Bandwidth Gap," a new report from Cahner's In-Stat Group, metro service providers will be making major investments in DWDM systems over the next five years. The research firm predicts that investments will jump from nearly $1 billion in 2000 to almost $8 billion by 2005.
The numbers out of IDC are more conservative; the research firm predicts worldwide revenues for metro DWDM will reach $2.3 billion by 2005, for a compound annual growth rate of 47%. The deployment of DWDM technology in the metro space would ease the problem of fiber exhaust as well as provide a significant increase in capacity. Also, because metro DWDM is an extension of long-haul DWDM, interconnections between the two can be done with relative ease.
Topic Networks has developed an IP-over-Ethernet solution that relies heavily on metro DWDM. "The idea is to give [service providers] flexibility in the metro network," explains Poranski. "DWDM is important. There's a need for fiber exhaust, recovering bandwidth, making more efficient use of those fibers, and packing those wavelengths and fibers as dense as possible."
To achieve widespread adoption in the metro space, however, DWDM technology must be adapted to the specific needs of the market. It is currently widely deployed in the long-haul segment, which generally comprises point-to-point connections. The metro space comprises numerous topologies, including meshes and rings, and long-haul DWDM is simply too rigid to meet these requirements. Metro DWDM is also very costly today, compared to other solutions.Several metro startups, including Mahi Networks, Metro-Optix, and Maple Networks, are banking on a multiservice provisioning platform, many of which integrate one or more of the aforementioned technologies. According to McEuen, these so-called "god" boxes, while initially attractive because of the breadth of functions they perform, will not hold up as a viable long-term solution. "I think as capacity and bandwidth issues become more and more of a problem, the god box isn't going to be able to handle it, because what it does is a little bit of everything," he asserts.Most metro vendors, regardless of the technology they back, are optimistic about the future. "There are problems and they have to be solved," asserts Poranski. "And that's opportunity." However, given the plethora of vendors in the space, that optimism is usually followed by a healthy dose of realism."In the long term, statistically, nine out of ten startups fail," admits Imregh. "They say when you flip a coin, you can have a run of 10 heads, but that doesn't change the laws of statistics. This year, we're running up on the 10 tails, I guess."Pioneer's McEuen agrees. "I think what's going to happen is that there will be more companies that just fold," he says. "There are some companies out there-and I'm thinking about the Ethernet service providers and the Ethernet equipment vendors. I think a lot of those guys have what it takes to make it, but there are just too many."Interestingly, says McEuen, the industry appears to have fallen into something of a pattern. The "hot area" has been moving and will continue to move closer and closer to the end user. "First, the problem arose in the long-haul, and that was addressed and handled," he explains. "Then it shifted closer to the metro, and now it's in the metro. I think in the future-5 to 10 years down the road-you're going to see a lot of people in the access market. That will be the hot area."
