U.S. wavelength-division multiplexing market to quadruple by the year 2000

With 1996 estimated sales of $80 million, the market for wavelength-division multiplexing equipment is projected to increase more than fourfold in the United States by the end of the decade despite tight competition from OC-192 time-division multiplexing equipment suppliers

mark lutkowitz

trans-formation inc.

The U.S. wavelength-division multiplexing (WDM) market is expected to grow strongly from 1996 to 2000, with equipment sales estimated to increase more than fourfold--from $80 million to $330 million, respectively (see figure). Based on this large WDM equipment market, Synchronous Optical Network (Sonet) time-division multiplexing (TDM) equipment suppliers are anticipated to undergo considerable pressure to make OC-192 (10-Gbit/sec) systems with necessary functionality and features available. As WDM and OC-192 equipment suppliers compete head-to-head for the 10-Gbit/sec and higher-rate markets mostly on price, profits are figured to be squeezed. Ultimately, only a few equipment suppliers are projected to succeed.

The following components were included in this WDM market study: transponders, fiber-optic multiplexers and demultiplexers, WDM cards in line terminal equipment, and power amplifiers at the transmit and receive ends. Line terminal equipment and line amplifiers were not covered.

Whenever a new technology is introduced into the telecommunications market, over-inflated projections of its growth in the industry frequently follow. However, some of the current optimism surrounding the WDM market appears to be justified for several reasons:

At 10-Gbit/sec rates, the telecommunications industry is pushing the physical limitations of today`s inexpensive, directly-modulated semiconductor technology. Higher rates for TDM systems might not occur until well into the next decade. The addition of coherent modulation technology and advanced lasers is likely to make the cost of a TDM system higher than 10 Gbits/sec prohibitive for a long time. While all sorts of new technology, such as solitons, is demonstrated in the laboratory, and while there may be unforeseen advancements, early commercial availability of cost-effective 40-Gbit/sec TDM solutions is problematic in the telecommunications market. This market tends to move at a snail`s pace.

AT&T is planning to deploy WDM technology across its huge, worldwide telecommunications network.

There is no end in sight to the impressive growth of bandwidth demand. The large interexchange carriers, which tend to install a fixed number of optical fibers, are projected to use WDM systems for increased capacity in their embedded networks.

Although it is useful to examine the preference for deploying WDM equipment on OC-48 (2.5-Gbit/sec) networks versus OC-192 TDM networks and equipment, ample room exists for both solutions in the U.S. telecommunications market. The buyer`s choice is expected to be made on a case-by-case basis based on cost. Entering into the cost equation are the anticipated growth in traffic, the nature of the network and business negotiations with vendors, among other factors.

Eventual widespread deployment of WDM equipment working in OC-192 networks is foreseen. Most OC-192 vendors are figured to provide standard wavelengths in the 1550-nm window, facilitating WDM deployment because the cost of a wavelength adapter is eliminated.

Although interexchange carriers and local telecommunications service providers are foreseen as the major WDM implementors, telecommunications market opportunities will be available to all owners of fiber-optic networks, including cable-TV operators, public utilities, state/local governments and private corporations.

Carrier demand

WDM technology will be a solution primarily for the interexchange carriers for several years. In contrast, the cost of WDM technology does not tend to prove economical for the short fiber-optic network distances used in traditional local telephone businesses. The break-even cost point for using WDM technology on a route that implements regenerators occurs at the second wavelength. The implementation of two wavelengths approximately equals the cost of that incurred with running two separate systems. When a third wavelength or more is implemented, a carrier will find WDM technology to be less expensive.

As the dominant interexchange car rier, AT&T has been trying to push the telecommunications industry toward WDM solutions. The carrier is expected to make a multibillion dollar investment in this technology for its global network; some of this investment includes the installation of other gear, such as line terminal equipment. AT&T prefers implementing WDM technology on mature OC-48 networks rather than going to emerging OC-192 networks.

Despite its stated commitment to multivendor implementations, AT&T will likely stay with Lucent Technologies` product line. This line, which includes an 8-channel WDM system and does not at present include OC-192 gear, is a primary driver of AT&T`s network deployment plans. However, AT&T has not precluded OC-192 deployment in the future.

Another major interexchange carrier, MCI, strongly believes that OC-192 WDM networks will be its primary systems, as opposed to WDM technology on an OC-48 network. Although its deployment of dispersion-shifted fiber is a factor in this decision, the carrier is making sure that WDM technology will work on its existing plant if proven cost-effective. Its primary Sonet TDM vendor, Northern Telecom, appears willing to agree to an unusually low price as part of a large-volume OC-192 contract.

Yet another interexchange carrier, Sprint, claims that current economic analyses demonstrate that WDM on an OC-48 network is more attractive than deployment of an OC-192 network. Its adherence to four-fiber ring solutions is seen as influential in not leaping to OC-192 technology at this time. Sprint has WDM systems in its network that are capable of reaching 16 channels and is testing a 32-channel system in its laboratories.

Still another inter exchange carrier, WorldCom, is figured to be a big buyer of WDM systems, especially since its original long-term plans call for keeping its OC-48 systems in the network as it deploys OC-192. The company also plans to double the capacity of its OC-192 systems with WDM technology.

Local exchange carriers

It will be problematic for the local exchange carriers to deploy additional fiber in large cities, and the comparatively lower cost of WDM systems may look attractive. Equipment suppliers might offer a product for shorter- distance networks in the future.

In general, large local exchange car riers do not currently need 10-Gbit/sec capability, whether WDM or TDM, for their traditional networks. Apparently, several regional Bell operating companies are having a difficult time filling up the capacity of their OC-48 networks. In addition, it will not be easy for the large local telephone companies to achieve a high level of success in new business areas, such as offering long-distance communications and switched video services. As a result, the major WDM system opportunity for local exchange carriers during the next several years is expected to be in traditional interoffice trunk-type transport services.

The most energetic local exchange carrier, Pacific Bell, is studying the possibility of installing 8- and 16-channel systems for its interoffice networks. The company might install 8-channel systems as early as 1998 or 1999. No other local exchange carrier seems committed to using more than 4-channel WDM systems at this time.

Other carriers that might make the leap to WDM technology sooner are Bell Atlantic and US West. There is a strong possibility of WDM deployment in Bell Atlantic`s interoffice environment, with installations probably starting in 1997. US West has some long interoffice routes and, as a result, might install WDM technology on a fairly large scale.

The other large local exchange carriers are either not beyond the laboratory test stage or not beyond the paper-review stage regarding WDM technology with more than two-channel capa city. Although BellSouth has deployed a small number of 2- and 4-channel systems, it is still evaluating them.

Winning suppliers

The big equipment supplier winners in the U.S. WDM market during the next several years are expected to be Ciena, Lucent Technologies, Northern Telecom and Pirelli. Alcatel may turn out to be a sleeper, while Ciena and Pirelli have already achieved leadership positions. Ciena leapfrogged the competition with the release of a 16-channel system. It targeted its efforts on getting Sprint`s business. By establishing itself with Sprint`s aggressive installation of WDM gear, Ciena has achieved credibility as a viable supplier. After Pirelli was slowed down in getting its 4-channel WDM system to work, the supplier is racing to keep up with Ciena on a 32-channel system. Ciena is now promoting a 40-channel system and it appears to offer a better network management scheme.

Pirelli can take advantage of its solid reputation as an optical amplifier and cable supplier to the U.S. market. The company has also been a large supplier to Sprint with 4- and 8-channel systems and will continue to benefit from Sprint`s position that it have a back-up supplier. Also, Pirelli is looking to differentiate itself by developing an active add/drop multiplexer to optically drop a signal out of the WDM stack along the route. Moreover, Pirelli is aggressive on low pricing.

Both Pirelli and Ciena will likely partner with one of the Sonet TDM players to gain some operations advantages and to facilitate eventual entry into the local exchange carrier market. In general, the local exchange carriers are hesitant to go with non-traditional manufacturers. Ciena`s total focus on the WDM product line will likely give it some advantage over Pirelli, but Pirelli has been an established fiber-optic equipment supplier to the U.S. market for some time.

Other WDM suppliers, however, have an advantage over Ciena and Pirelli because they offer line terminal equipment and can provide an integrated solution. These suppliers will have an edge in terms of cost and, potentially, reliability because half of the network elements (the transponders) are not needed. Apparently, some carriers have observed some performance degradation induced by transponders. If TDM suppliers of OC-48 and lower-rate equipment offer frequency-specific lasers, these devices should drive down the transponder market. Still, a huge installed base of OC-48 and OC-12 equipment already exists in installed networks and might potentially need transponders.

Despite the recent divestiture of AT&T, Lucent Technologies is expected to rack up heavy sales selling WDM systems to its ex-captive carrier. Indeed, Lucent might not even have to be competitive on price in its dealings with AT&T. So far, the supplier has made small penetration into other interexchange carrier markets with its WDM equipment. It remains to be seen if the other interexchange carriers will become comfortable with purchasing products from Lucent, given its continuing close relationship with AT&T. Moreover, Lucent is anticipated to be engaged with reorganization issues for a long time.

With the lion`s share of the non-AT&T interexchange OC-48 market and by taking a lead position with its OC-192 products, Northern Telecom is expected to piggyback sales of WDM equipment onto its installed base. It will also benefit from being a strategic partner with Pacific Bell. Northern Telecom`s official position is that it offers solutions that are needed, whether these are TDM or WDM products. However, in the inevitable market battle of its two product lines, TDM system sales are predicted to take priority. High-capacity TDM systems are its "bread and butter" in the transmission field. Also, the company has to recoup a larger investment in complex OC-192 development, as opposed to the relatively simple WDM technology.

Alcatel is likely to capture a small share of the WDM market while striving to make inroads with its elegantly designed TDM equipment. The supplier reportedly will be shipping a 16-channel system using fluoride-based optical amplifiers and is promoting a 32-channel system. Pacific Bell appears to be seriously looking at Alcatel`s 4-channel system. Yet Alcatel might be hampered by its small share of the OC-48 market and the recent financial problems of its French parent. Also, it is unclear how much market penetration the company can make with its OC-192 system. The supplier is looking to package its 10-Gbit/sec TDM system with its 16-channel WDM technology.

Japanese suppliers seem to be behind in developing WDM product lines. With its focus exclusively on the local exchange carrier market, Fujitsu does not appear to be in a hurry to release an OC-192 product or a WDM system. Fujitsu plans to offer a WDM system after its OC-192 system becomes available, which would double the capacity of its 10-Gbit/sec TDM system. The WDM system is expected to be integrated into the OC-192 products.

NEC has displayed an 8-channel WDM system that is upgradable to 16 wavelengths. However, NEC is not expected to become a large WDM player.

The prices for a 4- and 8-channel end-to-end WDM system are estimated to decline 49% and 43%, respectively, over the forecast period (1996 to 2000). During that time, the prices of 16- and 32-channel systems are projected to drop modestly by 35% and 19%, respectively. The sales of 32-channel systems is predicted to begin in 1998. The new and improved technology needed to operate 16- and 32-channel systems is expected to be the most important factor in the price drop of 4- and 8-channel systems, which use older technology.

In particular, the prices of WDM multiplexer/demultiplexer components--a small part of the overall cost of end-to-end WDM systems--look to be artificially high, given the non-complex technology. Suppliers want to recoup their equipment development investment. In three years, multiplexer prices could decline as much as 50%. Carriers are expected to demand much lower prices and play vendors against each other.

Prices to decrease

Prices for entire WDM end-to-end systems are figured to decrease for several reasons. With limitations on a sup plier`s ability to differentiate its products, other than possibly the network management scheme, pricing is projected to be a key marketing factor. Also, higher levels of production should reduce manufacturing cost, thereby allowing more flexibility to lower prices. Ultimately, prices should drop markedly as OC-192 systems become more cost-competitive.

Typically, in the electronics market, equipment becomes cheaper with more bandwidth. However, given the passive nature of WDM equipment where the technology remains constant, just the wavelength spacing is different. The common power supply and the management system are spread over more wavelengths. For example, on a 16-channel system, the second through the sixteenth channels tend to incur insignificant increments in price.

The maximum capacity on WDM systems is expected to remain at 32 channels during the forecast period. A network with 32 times the OC-48 bandwidth should be sufficient for the most bullish expectations of traffic requirements. Also, managing a stack of 40 OC-48 rings is more complicated than directing an OC-192 WDM network with one-fourth the number of stacked rings. u

Mark Lutkowit¥is president of Trans-Formation Inc., in Birmingham, AL.