"When the market comes back" is a turn of phrase heard often in telecoms circles. None of us knows exactly what that market will look like if and when it appears. But the market segment that has suffered most—long-haul—is the one that may benefit the most from the current slowdown.
While executives and product managers have been struggling for sales, the researchers who first helped fuel the surge in transmission capacity and distance are still at work improving long-haul technology. For this market to come back, we'll need to see considerably more traffic, of course. We'll also need to see real reductions in operating expenses, especially for 10-Gbit/sec systems. And that's the focus of the research effort.Linn Mollenauer, a researcher at Lucent Technologies Bell Labs (Holmdel, NJ), notes that for many scientists, the slowdown has meant a chance to let the science catch up with the performance demands from the carriers. "There really was a point, before the bubble burst, that service providers were beginning to buy the idea of low-cost, efficient service with all-optical technology," he says. "But the technology never really got off the ground before the disaster."
Before and after the disaster, his Lightwave Systems Research group has been working on ways to make the all-optical network appealing to carriers. The group has recently presented results in Optics Letters reflecting some advances in ultra-long-haul technology that have Mollenauer excited.*
The journal article describes a novel dispersion-managed soliton technique for 10 Gbits/sec, showing how transmission distances can be extended beyond 20,000 km without optical-electrical-optical regeneration. This distance is a considerable advance in scale and, when combined with all-optical switching, would allow carriers to span North America with robust self-correcting networks that lower operating expenses.
Solitons have been known since they were first observed in the form of solitary waves that traveled along 19th century canals. Mollenauer generated the first fibre-optic solitons experimentally in 1980, but interest was limited until optical amplification overcame the distance limitations imposed by electro-optical conversion. In 1988, Mollenauer's group extended transmission distance to 4,000 km using Raman amplification in a recirculating loop. And although WDM technology was the winner in terms of increasing transmission capacity, WDM used in conjunction with solitons has remained part of product planning strategy at several companies.
A classical optical soliton signal can travel great distances over fibre because the pulse spreading caused by chromatic dispersion is offset by the effects of self-phase modulation. More recently, dispersion-managed solitons have been developed to overcome the signal impairments found in classical solitons by splicing together two fibres having different dispersion properties to keep the total dispersion of the two close to zero.
Mollenauer's group has now taken a step beyond this technique by installing a periodic group delay dispersion-compensating module. The module is based on etalon technology from Avanex and compensates for only a small part of the transmission span. Other dispersion-compensating fibre is still used, but the result is to dramatically reduce the distance-limiting effect of jitter—the one significant nonlinear penalty not corrected by previous techniques. Jitter in signal pulses results from the collision between solitons in different wavelength channels.
This modification may seem like only a minor tweaking of system performance, but as Mollenauer explains, it means that dispersion compensation is independent of distance, economical, and compatible with DWDM systems operating at 10 Gbits/sec. In other words, it's ideal for continent-spanning all-optical networks.
More exotic distance-enhancing techniques such as differential phase-shift keying (DPSK) modulation may be the better dispersion-compensating approach for 40-Gbit/sec systems, says Mollenauer. But with his new periodic group delay dispersion management technique, standard on-off keying works just fine. The result is another enhancement to existing long-haul technology that should serve the industry well "when the market comes back."
*X. Wei, X. Liu, C. Xie, L.F. Mollenauer, Optics Letters, Vol. 28 (June 2003).