To 40 Gigabits and beyond
Speed and capacity of communications networks will be very much on the minds of attendees at this month’s OFC/NFOEC conference in San Diego, CA.
Speed and capacity of communications networks will be very much on the minds of attendees at this month’s OFC/NFOEC conference in San Diego, CA (Feb. 26-28). Nearly one terameter (1000 million kilometers) of optical fiber is now deployed around the globe according to plenary keynote speaker Herwig Kogelnik, adjunct photonics systems research vice president at Alcatel-Lucent’s Bell Labs. And with more than one billion Internet users continuing to demand newer and broader bandwidth services, lightwave research and development has advanced the capacity of long-haul fiber transmission systems by a factor of 100 since the WDM (wavelength-division multiplexing) revolution began just over 10 years ago, he says. His talk, “Perspectives on Optical Communications” is scheduled for Tuesday, Feb. 26.
One recent advance is transmission of single-channel data streams at 40 Gbit/s. Now that many of the technical hurdles have been overcome, serious commercial deployment started last year. That has been quickly followed by field trials at 100 Gbit/s, though many challenges remain (see page 79). Also seeing serious commercial deployment after many years of anticipation is fiber to the home/node/business (FTTx). Unexpectedly strong growth has pushed the overall transmission laser and transceiver market to nearly $1.59 billion in 2007, according to the Laser Focus World annual review and forecast, which is higher than was projected a year ago. Such growth will no doubt drive even stronger demand for speed and bandwidth (see page 59).
Meantime, work continues on many fronts to develop “all-optical” networks with more integrated, faster, and lower-cost components that can handle the ever increasing data rates. The conventional method of storing data in optical networks, for instance, typically involves an optical-electronic-optical conversion of signals and researchers are exploring a number of options to avoid such a conversion step (see page 34). And in another example, researchers are looking at vertical integration of optical elements in a CMOS-compatible process to enable “massive integration for high-density networks.” (see page 25).
Stephen G. Anderson
Associate Publisher/Editor in Chief