Fall of the Raman Empire?

As economic realities force carriers to slow the pace of new network buildouts, some vendors at last week's Optical Fiber Conference in Anaheim, Calif., exhibited technologies designed to let service providers squeeze high-speed capacity from their existing infrastructure.

Leading the way was OptiMight Communications, which unveiled the OMC 1600, a high-capacity core optical transport product designed to work over any type of fiber without electrical regenerators or Raman amplifiers. As part of OptiMight's full-spectrum WDM solution, the OMC 1600 lets carriers transport 1.6 Tb/s bi-directionally without regeneration for distances up to 4000 km (about 2500 miles).

While such performance is impressive in a greenfield environment, the key characteristic of OptiMight's system is that it can deliver these transmission rates over legacy fiber that lacks the quality characteristics necessary to support Raman amplification.

“For new networks, [Raman amplification] is fine — you have… new fiber,” said Ilya Fishman, OptiMight's president and chief technology officer. “But what are the old carriers going to do? Are they just going to die peacefully or what?”

OptiMight officials contend carriers can extract cutting-edge performance from low-quality fiber by deploying its solution. Using a proprietary formatting scheme that incorporates the company's full-spectrum technology to extend the signal's range, the OMC 1600 transmits light onto a system that includes erbium-doped fiber amplifiers that are usually placed in existing amplification sites.

To equalize the gain of all channels, OptiMight's system design includes signal conditioners that also allow for the adding and dropping of individual channels, according to Pat Montanino, OptiMight's vice president of sales.

This solution allows carriers to avoid the two largest drivers of long-haul costs — electrical regeneration in WDM systems and Raman amplifiers in DWDM, according to Clarel Thevenot, OptiMight's director of marketing. While its technology is proprietary, OptiMight's solution uses only “off-the-shelf” components. As a result, OptiMight can save a carrier as much as 60% on equipment and operational costs of existing solutions, he said (see figure).

OptiMight has completed lab trials with two large IXCs, including WorldCom, which submitted a paper with OptiMight at OFC detailing their lab test conducted on a 400 km recirculating loop with 40 10 Gb/s channels. Utilizing EDFAs every 80 km and signal conditioners at 400 km, successful transmissions were realized at distances from 4000 km to 6800 km, according to the report by David Chen, advisory engineer for WorldCom's optical and network division, and OptiMight engineers.

OptiMight expects to begin field trials soon and deploy the OMC 1600 during the fourth quarter of this year. The box will be upgradeable to a 3.2 Tb/s system via a card replacement in 2002. OptiMight plans to offer 6.4 Tb/s capability — an upgrade probably requiring a new box — during the latter half of 2003.

While OptiMight's solution generated considerable interest during OFC, the company's reluctance to disclose how its technology works fostered some skepticism among industry observers, especially before the WorldCom lab tests were disclosed.

“I don't think you can afford to simply dismiss anything — nobody knows who's got the answer,” said Optisphere CEO Jost Spielvogel. “But, like anyone, they have to prove it works.”

Using more traditional technology, Hitachi Telecom also announced a solution that utilizes a carrier's existing fiber. Norlight Telecommunications, a private regional carrier in the midwestern U.S., announced it will pay Hitachi Telecom $40 million to upgrade its network from OC-48 rings to performance levels of 10 Gb/s.

“We think there's a good-sized market with these legacy networks,” said David Foote, Hitachi Telecom's director of optical products and technology. “There are a lot of carriers out there that don't have the newer fiber.”