Raman Risks Emerge
The need to avoid this drawback –- that distributed Raman pumping might actually damage fiber and create safety hazards in some circumstances –- explains why WorldCom Inc. (Nasdaq: WCOM) has conducted lab trials of equipment from OptiMight Communications Inc., which aims to support ultra-long-haul transmission without using Raman technology (see OptiMight Details Long-Haul Box).
Here’s the score. Distributed Raman works by pumping lots of light into the outer cladding of optical fiber in a way that boosts the strength of the signal running through the core.
Worldcom has found that things can go wrong when fibers carrying this extra load of light are disconnected from patch panels (the boards used to link one fiber to another), as they're likely to be when engineers implement network changes.
“If the connector has some dirt on it, you can damage it. If you’re not careful, the tip can melt,” says David Chen, advisory engineer in the optical and data network division of WorldCom Inc. (Nasdaq: WCOM). “We see a potential maintenance issue with Raman amplification,” he says.
”It’s not only a technical issue. It’s a safety problem as well,” notes Philippe A. Perrier, director of photonic subsystems engineering at Xtera Communications Inc. Such powerful light could injure the engineers working on patch panels, he warns.
Perrier doubts whether light would be powerful enough to melt the fiber but says that it can carbonize the tip if the connector is dirty. The tip would then require regrinding and polishing.
Xtera, by the way, uses “discrete” rather than distributed Raman amplification to boost signals in its ultra-long-haul S-band amplifiers (see Xtera's $110M Surprise). This means that pumping occurs at intervals, over limited lengths of fiber, well away from patch panels.
Nortel Networks Corp. (NYSE/Toronto: NT), which makes distributed Raman equipment, acknowledges that fiber could be damaged if couplers are dirty. However, it plays down the potential safety hazard. Nortel incorporates technology that automatically shuts down lasers in the event of problems, a spokesperson says.
These concerns led Worldcom into conducting trials of Optimight equipment, which supports ultra-long-haul transmission without using a Raman approach. In the trial, Worldcom and Optimight successfully transmitted 400 Gbit/s (40 channels at 10 Gbit/s a channel) for distances of up to 6,800 kilometers.
Here’s where the OFC connection comes in. Chen and a bunch of engineers from Optimight submitted a report on their trial for inclusion in the OFC’s prestigious “post deadline papers,” to be presented at the conference tomorrow (Thursday). Getting a post deadline paper accepted is considered a big deal by engineers. It equates to recognition of world-class research.
To cut a long story short, Chen’s paper was one of the many papers that didn’t get picked for tomorrow's session. Something like 900 submissions were made and only 39 made it through the OFC’s selection process, according to Chen. His paper is posted on http://www.optimight.com/news/03202001.html
One reason why Chen’s paper may not have made it is that it describes the trial without mentioning the real reason for conducting it in the first place -- to examine ways of avoiding potentially significant maintenance problems.
Light Reading has been told unofficially that Chen’s paper wasn’t accepted because Worldcom’s trial isn’t a record. Even longer distance transmissions without Raman amplification have been reported in the past.
This is disputed by Optimight, which says that Worldcom wouldn’t have gone ahead with the trial if the technology already existed. The implication is that the folk selecting the OFC’s post deadline papers wanted to avoid the whole issue, possibly because of vested interests. Chen has declined to comment, saying he doesn’t want to get involved in vendor politics.
-- Peter Heywood, international editor, Light Reading http://www.lightreading.com