If the hero experiments that dominate conference post-deadline papers were an exact predictor of market trends, then we could soon expect to see carriers deploy all-optical 160-Gbit/sec systems. Of course, that's not the purpose of the papers or the experiments they describe, which often push a technology well beyond expected operating conditions, so it can be refined and proven-in for real-world applications.
Perhaps one-quarter of the 47 post-deadline papers at March's OFC could be called hero experiments—that is, experiments in which research teams test a system's performance and capacity by pushing ingredients like distance, speed, modulation format, and channel spacing. Most take place in the lab, but one real-world hero experiment was performed by researchers at Tyco Communications (PD27). Their field trial ran 128 channels at 12.3 Gbits/sec over a single dispersion-flattened fibre between Hillsboro, OR, and Toyohashi, Japan, a distance of 8,991 km. That wasn't the most heroic of experiments, but it was impressive because it demonstrated terabit-per-second-per-fibre capacity on an installed undersea system.
All-optical regeneration has been on the industry wish list for many years, since it would eliminate the expensive optical-electrical-optical (OEO) conversion now necessary to re-amplify, reshape, and retime (3R) optical signals. An all-optical regenerator would extend transmission distances and, if it had wavelength-shifting capability, allow carriers to quickly reconfigure their networks, including optimising use of the many different types of fibre in their networks.
There were three post-deadline papers at OFC on all-optical regeneration. The approaches varied from a nonlinear fibre-based 2R regenerator and dispersion compensator operating at 40 Gbits/sec from Corning (PD05) to an Alcatel 3R regenerator based on cascaded semiconductor optical amplifiers, also at 40 Gbits/sec (PD15). A Fujitsu/Heinrich-Hertz Institute team created a fibre-based wavelength-shift-free 3R regenerating repeater operating at 160 Gbits/sec (PD16).
As for post-deadline papers foreshadowing commercial products, on the OFC exhibit floor, Lightbit launched the first commercial all-optical 2R regenerator. Lightbit chief executive Larry Marshall said carriers and OEMs are quite interested in the product for next-generation 10-Gbit/sec systems. He acknowledged that 3R regeneration is better than 2R, but argued that the third R—retiming—adds substantial expense and is not necessary for distances in North America. The regenerator is based on a buried waveguide in a magnesium oxide-doped lithium niobate wafer.
The line between such commercial products and "research" is a bit blurry in the post-deadline world. In the case of researchers from Ahura, the fact that a post-deadline paper was accepted at OFC motivated them to come out of stealth mode and discuss their high-power incoherent Raman pump module (PD47). Marketing director Eric Schmidt says the company saw the opportunity when the post-deadline committee selected them on the basis of their use of a broadband source to replace a pump laser and the fact that their distributed Raman amplifier fits in a butterfly package (see Photo).
The selection process for post-deadline papers happens quickly but methodically, according to OFC technical chair Tom Afferton of AT&T. He and co-chair Doug Baney from Agilent Labs led the nine technical committees that selected all conference and post-deadline papers. Afferton says the acceptance rate for OFC post-deadline papers has remained the same over the years at about 40%.
I wondered if the committees had a natural predilection for hero experiments. Afferton assured me that the selection was based strictly on the papers submitted. Hero experiments—rather than more theoretical work or research requiring greater explanation—simply lend themselves to the short, breaking news captured in post-deadline papers.
The first criterion for post-deadline papers in all fields is that they advance the science or technology. Based on the evidence in the telecom world, commercial potential is a reasonable second.