The proposed one year project will explore high performance architectures applicable to realizing optical WAN switches. Some of the major drawbacks of currently proposed high speed "optical" switches are the need for rapidly tunable transmitters or receivers requiring fast control for contention resolution and for transmitter-receiver coordination. Alternatively, current architectures based on the use of fixed transmitters and receivers, as in a Shuffle, result in good performance only for homogeneous traffic patterns and do not smoothly adapt to component failures. The principle of the proposed solution is the use of wavelength or frequency routing, in a way which trades the high bandwidth of these technologies for simplifying the processing and hardware switch requirements. the proposed solution integrates time oriented control with wavelength division multiplexing (WDM) targeted to overcome the above limitations. The objective of the one year effort will be to explore the major issues that restrict WDM/TDM (time division multiplexing) operations and propose a preliminary switch architecture. The operation of the proposed system will be evaluated through simulation to provide a quantitative measure for evaluating the future potential of the proposed approach.
