Implementation of Dynamic Bandwidth Re-allocation in Optical Interconnects using Microring Resonators

Dynamic bandwidth re-allocation (DBR) technique balances traffic by re-allocating bandwidth from under utilized links to over utilized links. This results in significant improvement in overall throughput and latency. In previous study, passive optical devices, namely arrayed waveguide gratings (AWGs) and couplers were used to implement DBR [2]. Although the performance was significantly improved, the design was cost prohibitive since it required n² number of lasers per board (where n is the number of transmitters per board). In this paper we propose the implementation of DBR using silicon on insulator (SOI) based microring resonators. We call this technique row-column switch implementation. The proposed active implementation reduces the number of required lasers by a factor of n without any degradation in performance. Analytical and simulation studies were conducted to compare the passive implementation of DBR with the proposed active approach. This comparison included area, power consumption, throughput, power loss (dB), power-delay product and area-delay product. Results show that the proposed active switch provides throughput and latency similar to the passive implementation of DBR while dramatically improving cost by a factor of n. There is a slight increase in power consumption (0.4% for the worst case traffic) using the active switch implementation.