Reducing the cost of fiber-to-the-home: Reflective MEMS modulators at the end user impress the upstream data string on a diverted portion of downstream power, eliminating the need to house a costly transceiver at each home.
photonicsonline.com{B7AD3914-4B8F-11D4-8C54-009027DE0829}&Bucket=Technical+News&VNETCOOKIE=NO
With long-haul networks moving toward an all-optical system and demand for high-bandwidth applications growing, optical components companies say FTTH is the next logical step in bringing consumers high-speed connections, up to 10 Gb/s. Already pilot programs are being run and cable is being laid to make the service possible, but analysts say how fast these dreams become reality will depend on both the costs of installing the systems and how much demand there is for the service.
“I think things will happen over a period of time, but not in the next few years,” said Claude Romans, director for access networks at market research firm RHK (South San Francisco). “I think it’s getting close, but I think you have to step back and say ‘why do I need this?’”
Jeff Montgomery, chairman of market researcher ElectroniCast (San Mateo, CA), agrees that equipment has to become cheaper and a profit model has to be developed. But the market for FTTH is real and growing, he said, projecting that connections will be available to 190,000 homes this year, with the figure rising to 1.59 million in 2004.
Lose the laser
One issue of FTTH is the state of the metro and access infrastructure—FTTH implies a fiber link stretching all the way from the nearest trunk line. Assuming that issue is surmounted, however, a second major concern is cost. A bidirectional FTTH link requires the presence of a transceiver at each home, driving up system cost, as well as increasing the complexity of installation. Wayne Knox and colleagues at Bell Laboratories (Murray Hill, NJ) are exploring a very different approach, however, one that would eliminate the need for an upstream transmitter.
Knox's system takes advantage of the broad bandwidth afforded by a femtosecond laser to generate as many as 4000 closely-spaced wavelengths between 1.5 and 1.6 µm. In practice, the laser would be located at a controlled environment in the central office. At the home, a splitter divides the optical power—a portion passes to the detector, and a portion passes down a fiber length to reflect off of a micro-electro-mechanical systems (MEMS) based modulator, which impresses the upstream signal on the beam.
The scheme irresistibly brings to mind the image of a Boy Scout signaling his troop in Morse code by reflecting sunlight off of a mirror, but Knox and his crew have demonstrated feasibility, sending signals up to 30 km out and back from the source (central office). Using a MEMS-based modulator makes the proposal economically viable. "The whole idea with MEMS is that it is scalable to very low cost," noted Knox.
High data rate and flexibility
The major cost lies in the fiber deployment. According to Knox, in greenfield locations, deployment of FTTH is cost competitive with twisted pair deployment. "Our system provides T-3 [data rates] to scalably large groups of customers," he said.
With so many wavelengths available, the system could drop 100 channels at each node, perhaps becoming as granular as one wavelength per house, he added, contrasting the approach with that of cable modem, in which speed drops as the number of users rises. "You'd never have to give 45 Mb/s to anyone, because it's your own wavelength. Shared data networks will eventually become as antiquated as a party line."
The approach would also provide carriers with the flexibility to address demands of varied customers. A customer with high capacity demand could swap the MEMS modulator with a distributed feedback (DFB) laser operating at 10 Gb/s.
About the author…
Neil Savage is a freelance writer based in Massachusetts |
