OFC gets up to speed for next generation networks

OFC gets up to speed for next generation networks
The technical program for this week's Optical Fiber Communications (OFC) Conference in Anaheim, Calif., reveals researchers worldwide mounting a broad-based attack on the problems of designing ultrahigh-speed optical networks. At the same time, several OFC presentations will tell of new records in fiber-optic transmission rates.

While commercial systems have finally arrived at the 10-Gbit/second level, practical demonstrations of 40-Gbit/s components and networks are cropping up everywhere. And one OFC paper describes a successful field trial of a 160-Gbit/s transmission by researchers at the Heinrich-Hertz-Institut (Berlin). The trial was conducted several times on existing fiber connections among three German cities, with a perfect score of zero errors every time.

Several of this week's In Focus contributors are presenting their work at the conference, disclosing highly novel approaches to optical-component design. One example, explained by researchers from Lucent Technologies Bell Laboratories (Murray Hill, N.J.), is a microfluidic approach to building tunable optical components inside the fiber itself.

The system is fairly simple. A segment of fiber is built with six hollow channels arranged symmetrically around the core and running the length of the segment. Inserting fluid into air-silica fibers makes it possible to modulate the optical signal in the core. The fluids shuttle down the fiber by attaching an external, electrically activated heater, resulting in tunable devices such as wavelength-selective attenuators and switchable broadband attenuators. According to Lucent, this innovation represents an important new direction for photonics research with the potential to provide essential technologies for next-generation optical networks

Meanwhile, Payam Rabiei and William H. Steier, researchers at the University of Southern California (Los Angeles), are using inexpensive polymers to make very small optical components on a silicon chip. In their contribution, the scientists explain how they constructed a microscopic polymer version of an electro-optic modulator. The polymer micro-ring resonators vary in diameter from 40 to 400 microns. According to Rabiei and Steier, it is also possible to incorporate them into standard CMOS electronic circuits.

Presently, companies are planning to install 40-Gbit/s electrical time-domain multiplexed (ETDM) systems. But for the next-generation 160-Gbit/s bit rates, electrical signal processing does not work. The Heinrich-Hertz-Institut project, conducted in tandem with a group at Lucent Technologies Nuernberg, has managed to leapfrog the 40-Gbit/s generation with an optical TDM system. Data generated at 10 Gbits/s was multiplexed up to 160 Gbits/s and launched into installed fiber maintained by Deutsche Telekom. The signal traveled through 116 km of standard single-mode fiber stretching among three cities. It was then demultiplexed optically to 40 Gbits/s and further demultiplexed with a Lucent optoelectronic demultiplexer. The researchers report zero error transmission.

Polarization-mode dispersion (PMD) has been considered a significant barrier to achieving single-channel data rates at 10 Gbits/s and beyond in optical communication systems. Carsten J. Videcrantz, product-marketing manager, and Florin Rosca, optical engineer, at Mintera Corp. discuss how optical and electronic service providers may soon be able to harvest the benefits of 40G in terms of lower cost and faster service provisioning. Specifically, they dispel the myth that chromatic dispersion at 40 Gbits/s will sink cost-effective implementations. The reality, Videcrantz and Rosca maintain, is that chromatic dispersion reduces nonlinear effects in the fiber and as such enables 40G dense wavelength-division multiplexing (DWDM) over long distances when combined with well-known in-line dispersion-compensating fiber.

Researchers at the Department of Computer Sciences and Electrical Engineering at the University of Maryland (Baltimore) describe a design that tackles the more difficult problem polarization-mode dispersion poses. Called polarization diversity receiver, the design uses simple fixed optics and some electronics to control PMD at high speeds.

On the transmission front, researchers at Agere Systems Inc. say they have set a record in transmission capacity: 3.2 terabits/s over a 1,000-km fiber using DWDM. The transmission consisted of 40 channels operating at 80 Gbits/s each. And Mitsubishi in Japan says it has achieved a transoceanic-class transmission-rate record of 1.3 Tbits/s over 8,400 km on a single fiber.

Mitsubishi's research group used 65 channels at a bit rate of 20 Gbits/s, which they say offers the optimum solution in terms of potential for lower cost and reliable transmission compared with bit rates of 10 and 40 Gbits/s. The Mitsubishi approach may open the door for next-generation submarine cable systems and terrestrial systems that would connect Japan, the rest of Asia, the United States and Europe.

See related chart