THE DEMAND FOR DWDM COMPONENTS AND SUBSYSTEMS IS STRONG, ALTHOUGH NOT AS STRONG AS SOME FORECASTS PROJECT. COMPOUND ANNUAL GROWTH OVER THE NEXT FIVE YEARS SHOULD BE 30%, BUT SUCCESSFUL MANUFACTURERS MUST CHOOSE SEGMENTS THAT TARGET INDUSTRY DEMANDS.Lisa Huff
The components industry for dense wavelength-division multiplexing (DWDM) is in a period of flux. After a boom of spending by service providers and carriers in 1999 and 2000, the market for optical networking equipment has started to slow. We have always believed that predictions of 500% growth over the next three years were unrealistic, yet at this point we do not believe there is cause for alarm. Although the economy is slowing and companies such as Nortel Networks, Lucent Technologies, and JDS Uniphase are announcing layoffs, every day sees another optical components startup company announcing a new product or multiple millions of dollars of investment financing. This trend indicates that, even though existing business is slowing down, carriers, equipment manufacturers, and component suppliers are still planning for the future.
Over the past two years equipment manufacturers have experienced shortages of optical components unequaled in history. Some components had lead times as long as 12 months. This situation has corrected itself by components suppliers adding capacity as quickly as possible and by the slowing of the economy. Most components suppliers are now finding themselves with excess capacity and are having to find new customers to help utilize it.
Specific to the DWDM market, component suppliers have seen a shift in demand from the long-haul and ultralong-haul markets to the metropolitan market. This trend has caused headaches for the suppliers because requirements for components for the metro market can be vastly different from those of long haul and ultralong haul. For example, the essential long-haul market requirement has always been bandwidth, whereas the metro market requirement is connectivity—these requirements translate into very different needs for optical components and subsystems.
We see a tremendous opportunity for suppliers in the optical components and subsystems arena, if they understand the overall picture of the optical networking market. Current players and new entrants into the DWDM components and subsystems market must consider some complex realities:
- Development of next-generation components and subsystems
- Changes in the supply chain for such products
- Impact of new optical materials and processes
- Growth of metro, access, and enterprise markets; on-going support of ultralong-haul and long-haul markets
- How components development affects network design
- Tighter venture capital funding in an economic slowdown
- Lower capital expenditure by carriers and service providers in an economic slowdown
There are four major network configurations for DWDM systems: point-to-point, ring, mesh, or star. Most configurations deployed today are either point-to-point or ring topologies. Most experts would agree that mesh configurations have many advantages over ring configurations but, taking the installed infrastructure of SONET systems into consideration, rings seem the most prudent deployment for the time being. Eventually the entire long-haul network will become a mesh network.
Service providers such as AT&T, MCI Worldcom, and Sprint have been deploying DWDM systems for more than five years. Their customers are now insisting on "bandwidth on demand services." To provide more bandwidth, DWDM systems have increased in data rates from 622 Mbit/s (OC-12) to 10 Gbit/s (OC-192). Systems that will operate at 40 Gbit/s (OC-768) are being developed, but new components technologies are needed to overcome the issues of polarization- and chromatic-mode dispersion associated with systems running at such a high data rate.
Tighter channel spacing is also helping to provide more bandwidth. Channel spacing has gone from 400 GHz in the original systems down to as low as 1 GHz being developed today. Channel spacing at 1 GHz will not be needed immediately and 40 Gbit/s data rates will not be used with this spacing, but components and subsystems suppliers are looking at every avenue available to help carriers increase services to their customers.
Of course, none of these advances in service provisioning would be possible without the research and development of component and subsystem suppliers. Some of the most important advancements in DWDM components include tighter channel spacing, tunable devices, integrated components, and expansion into the L- and S-bands. Technologies enabling these components include fiber Bragg gratings, arrayed waveguide gratings (AWGs), fused fiber, planar lightwave circuits (PLCs; also known as photonic integrated circuits), vertical cavity surface emitting lasers (VCSELs), and Raman amplification. Companies on the cutting edge of these technologies include the traditional telecom players as well as many new startups (see table).
Most important to the future of DWDM systems are tunable components. When tunable lasers replace existing one-wavelength lasers, costs to produce and maintain DWDM systems will be drastically reduced. Instead of having a spare laser for every wavelength being transmitted, service providers will only need to have one for every four or eight wavelengths (depending on how widely tunable the laser is). The other benefit realized by tunable lasers is dynamic services. Components manufacturers are designing the tunable lasers to be programmable, allowing them to be changed remotely instead of by a site visit.
Another important technological advance that will revolutionize DWDM systems is integrated components. Wavelength lockers, modulators, isolators, and attenuators may all be integrated within the transmitter instead of designed as discrete components. This development could reduce the size of equipment since all of these functions would be incorporated in an all-inclusive integrated module or subsystem.
Some of this integration is already taking place in tunable transmitters—Bandwidth9 and Agility Communications have tunable transmitters that incorporate wavelength lockers. Most component suppliers are including modulators in their transmitters. Once these devices become readily available, equipment manufacturers may have to re-think the whole network architecture to take advantage of the new integrated functions.
Other advances in optical amplification will be important to track as amplifiers become powerful enough to allow fiber spans of over 1000 km without regeneration. For this purpose, Raman technology will not only help with long-haul amplification but will allow expansion of wavelengths into the S-band (1480 to 1525 nm). Xtera Communications, utilizing devices developed by Oplink Communications, has just released subsystems that take advantage of Raman technology.
The great majority of research and development resources are being expended on devices to address metropolitan area networks. Optical components suppliers find they must redesign or develop new products to address this market since the components developed for the long-haul market are too expensive. This factor is why coarse WDM was originally deployed in the metro market—it was more cost effective. Now that the long-haul networks are deploying DWDM and providing higher data rates, the bottleneck has moved to the metro network. New cost-effective devices are being developed to increase connectivity. For example, Essex is developing hyperfine WDM technology that will allow channel spacing as close as 1 GHz in a device with a cost target of $500.
DWDM COMPONENTS MARKET
Some analysts have predicted that the DWDM component market will grow 500% by the year 2003. We are skeptical of this number and CIR is forecasting a much slower growth rate. Our methodology is based on understanding the optical networking business through discussions with service providers, then interviewing equipment manufacturers, and finally talking to component and subsystem suppliers. Based on this approach, we see a total compound average growth rate over the next five years of 30% (see figure).
Although the economy has taken a turn for the worse, there is still plenty of opportunity for optical component suppliers. The shakeout that is happening in the service provider, equipment manufacturer, and optical component supplier industries is healthy. Too many companies were over-valued for too long, so the correction is actually refreshing.
Opportunities abound for optical components suppliers if they stay focused on the goals of improving cost effectiveness in each part of the networks—ultralong haul, long haul, and metro—and developing the tools that carriers need to improve services through innovation and integration.
Lisa Huff is a senior optical components analyst with Communications Industry Researchers, P.O. Box 5387, Charlottesville, VA 22905 . She can be reached at 717-795-8355 or firstname.lastname@example.org.