100-km Negative-Dispersion Fiber Carries 10 Gb/s

Richard Gaughan

To satisfy consumer demand for higher bandwidth in metropolitan-area networks, wavelength division multiplexing (WDM) systems must include economical, application-optimized components. Researchers have reported a new fiber with negative dispersion from 1280 to 1620 nm that has the ability to carry 32 signals at 10 Gb/s over 100 km.

Directly modulated distributed feedback lasers are cost-effective sources of 1550-nm light that offer 10-Gb/s modulation rates and outputs of 5 to 10 mW, but they display positive-frequency chirp, in which the leading edge of a pulse is blue-shifted while the trailing edge shifts to the red. If the zero-dispersion wavelength of single-mode dispersion-shifted fiber is matched to the laser, long-distance single-channel transmission is possible over the span. When more lasers are added, however, cross-channel nonlinear effects limit their reach. On the other hand, fibers with sufficient positive dispersion at the 1550-nm wavelength cause the bit patterns to overlap, reducing the effective transmission length to only 10 km.

9-dB Q-factor over 100 km

Scientists at Corning Inc.'s Photonics Research and Test Center in Somerset, N.J., measured the transmission characteristics of the company's MetroCor negative-dispersion fiber with distributed feedback lasers from different commercial suppliers. At 10-Gb/s rates, their initial tests with an extinction ratio of 8 dB yielded error rates of better than 10^-15 through a 75-km fiber, corresponding to a Q-factor of 9 dB. With a 5-dB extinction ratio, they found the same Q-factor over 100 km of fiber.

Ioannis Tomkos, a researcher working with Richard Vodhanel's group at the center, extended the testing to a 32-channel WDM architecture with up to four nodes over 100 km of the fiber. The 10-Gb/s channels, which were separated by 200 GHz, each demonstrated a Q-factor better than 9 dB.

According to Tomkos, the 320-Gb/s aggregate bandwidth system demonstrates the feasibility of cost-effective high-capacity networks serving the majority of European metro needs.

The team plans to evaluate the fiber's performance with other transmitter types, such as wavelength-tunable vertical-cavity surface-emitting lasers.