Sonet equipment market looms large on the horizon
MARK LUTKOWITZ
The U.S. market for synchronous optical network OC-192 (10-gigabit-per-second) terminals, not including regenerators, is projected to surge from more than $5 million in 1996 to approximately $62 million in 1999--a jump of more than 1100%. The main drivers for optical carrier, level 192 hardware purchases are expected to be similar to those that previously influenced the move to OC-48 (2.5 Gbits/sec)--higher capacity over fiber-optic networks to support additional subscriber services. The major buyers of OC-192 hardware are projected to be the long-distance carriers.
Throughout this year, MCI, Sprint and Wiltel have expressed considerable interest in deploying OC-192 systems for increased capacity. In fact, these three long-distance carriers and AT&T are likely to purchase two-thirds of the total OC-192 terminal shipments in the United States during the next five years.
As a result of expanded network needs, MCI would have implemented OC-192 hardware during 1994 if such equipment were available. The carrier currently delivers most of its traffic on OC-48 equipment. However, this capacity is completely committed. For that reason, MCI will be a primary installer of OC-192 equipment. Company plans also include the use of wavelength-division multiplexing on OC-192 to achieve 40-Gbit/sec capacity.
Despite its need for OC-192 capacity, Sprint may in some cases prefer OC-48 systems with four-window wavelength-division multiplexing until more attractive price projections are received from OC-192 system suppliers.
Wiltel does not plan to initially replace its OC-48 equipment with OC-192 terminals. For the long term, OC-192 implementations systems will most likely coexist with OC-48 equipment.
Unlike its three long-distance competitors, AT&T seems to be in no hurry to install OC-192 systems. In fact, it is leaning toward wavelength-division multiplexing technology to increase capacity. Apparently, AT&T does not have the service capacity problems that its competitors do. However, the carrier should be a major buyer of OC-192 systems during the next five years because AT&T Network Systems is expected to be a major OC-192 system supplier.
As for local exchange carriers, Pacific Bell is likely to be aggressive in purchasing OC-192 systems to meet its increasing demand for video and data services. Other LECs, including Bell Atlantic, US West and Southwestern Bell (re named SBC Communications Inc.), are analyzing their future OC-192 applications. These telephone companies are expected to deploy OC-192 systems in the interoffice portions of their networks, mainly in large metropolitan areas. Executives at these telephone companies say they do not envision the use of OC-192 systems in the local loop.
Some suppliers have prototype OC-192 equipment available. Trialing of commercial equipment could start by early 1996, and large deployment of OC-192 systems is forecast to begin in 1997.
An incumbent Sonet equipment manufacturer--one that is currently serving a particular carrier--is strongly positioned to sell OC-192 systems to that carrier. The system must be priced right, provide the necessary features and be reliable. Equipment suppliers do not yet provide plug-and-play interoperability. Consequently, carriers are sticking to their incumbent suppliers to reduce the problems they have with operations systems and networking management. In addition, they avoid hassles with new training and documentation. Through their long, strategic relationship with a manufacturer, carriers are able to customize requirements faster and at lower prices (see Lightwave Telecommunications Fiber Deployment Supplement, January 1995, page 18).
From an incumbency standpoint, Fujitsu Network Transmission Systems Inc. is a leading supplier and delivers Sonet transport equipment to nine large U.S. carriers. It also has strategic Sonet partnerships with SBC Communications, Nynex and MCI.
AT&T`s strength is the same for all its public network products--breadth of product line and strong ties to the LECs. Northern Telecom Inc.`s product strength lies in its high-capacity Sonet transport systems, in particular, the OC-48 system. Its strategic partnership with Pacific Bell, a potential large buyer of OC-192 systems, has great sales potential.
Alcatel Network Systems Inc. is expected to be a strong competitor in the OC-192 market. It is already a major Sonet digital crossconnect supplier to the U.S. market. NEC America Inc. will be an aggressive supplier in producing viable OC-192 systems early. It supplies Sonet transport terminals to MCI, Sprint, MFS and US West.
Hitachi America Telecommunications Division is a new supplier to the U.S. public telecommunications market. Although Hitachi announced an OC-192 terminal at Supercomm `95 in March, it will need to be aggressive on delivery, price and customer support.
Technological and cost hurdles
Most large fiber-optic networks in the United States can work with OC-192 equipment. However, going from OC-48 to OC-192 networks means the quadrupled bit rates reach the dispersion limit of fiber in one-fourth the distance. Consequently, repeaters must be added. Solutions to this problem include improved lasers, the use of coherent modulation or the installation of dispersion compensating fiber.
Systems that utilize elaborate dispersion compensation are expected to be complex to maintain because they have to be fine-tuned, perhaps for each installation. During the life of the system, recompensation may also be needed because some fiber characteristics might change. Consequently, additional costs to compensate for dispersion, either in terms of higher OC-192 hardware expense or higher cost for network upgrades, could hamper growth in the OC-192 market during the next five years.
Another potential problem involves signal losses or reflections at the higher OC-192 bit rates because of bad connections in the network. Problems could arise because of mechanical splices, inferior fusion splices and older connector types.
Price trends
The OC-192 products are in their early technology phase. It is, therefore, premature for suppliers to fix pricing before they finalize product features and capabilities. However, carriers will probably not accept an OC-192 price that is more than two times the price of an OC-48 product. In 1996, OC-48 product prices are predicted to average approximately $115,000.
A wide range of pricing that depends on several factors is also forecast, including whether a variety of interfaces on the low-speed side is needed, a terminal mode or a ring mode is used, and whether termination into an ATM switch is required. Moreover, regional price differences should develop--even with the same supplier--because of marketing reasons.
Users are expecting the price decline for OC-192 products to follow a curve similar to that previously demonstrated by OC-48, OC-12 and OC-3. The extent of price drops is expected to depend heavily on volume purchases.
Fewer steep price declines compared with past equipment generations are predicted. The price decline during the next five years is projected to be a modest 15%. During that time, only four or five carriers are expected to make substantial purchases of OC-192 equipment. The 15% price decline should be somewhat driven by competition among Sonet suppliers, as well as the cyclical pattern of Sonet equipment price declines. q
Mark Lutkowit¥is president at Trans-formation Inc. in Birmingham, AL.