This paper presents a way of building modular systems with a powerful optical interconnection network. Each module, placed on a Printed Circuit Board (PCB), has a generic optical communication interface with a simple electronic router. Together with optical switching using micro-electromechanical system (MEMS) technology, packet switching over reconfigurable topologies is possible. The interconnection system gives the possibility to integrate electronics with optics without changing existing PCB technology. Great interest from industry is therefore expected and the cost advantages are several: reuse of module designs, module upgrades without changing the PCB, low-cost conventional PCB technology, etc. In the version described in this paper, the interconnection system has 48 bidirectional optical channels for intra-PCB communication on each board. For inter-PCB communication, a backplane with 192 bidirectional optical channels supports communication between twelve PCBs. With 2.5 Gbit/s per optical channel in each direction, the aggregated intra-PCB bit rate is 120 Gbit/s full duplex (on each PCB) while the aggregated inter-PCB bit rate is 480 Gbit/s full duplex. A case study shows the feasibility of the interconnection system in a parallel processing system for radar signal processing.

©2002 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

An optoelectronic router with a shuffle exchange network is presented and enhanced by adding micro-optical-electrical mechanical systems (MOEMS) in the network to add the ability to reconfigure the shuffle network. The MOEMS described here are fully connected any-to-any crossbar switches.The added reconfigurability gives the opportunity to adapt for different common application characteristics. Two representative application models are described, where the first has symmetric properties and the second has asymmetric properties. The router system is simulated with the specified applications and an analysis of the results is carried out. By using MOEMS in the optical network, and thus reconfigurability, over 50% increased throughput performance and decreased average packet delay is obtained for the given application. Network congestions are shown to be avoided throughout the system if reconfigurability is used.

Application specific architectures are highly desirable in embedded parallel computing systems at the same time as designers strive for using one embedded parallel computing platform for several applications. If this can be achieved the cost can be decreased in comparison to using several different embedded parallel computing systems. This paper presents a novel approach of running several high-performance applications concurrently on one single parallel computing system. By using a reconfigurable backplane interconnection system the applications can be run efficiently with high network flexibility since the interconnect network can be adapted to fit the application that is being processed for the moment. More precisely, this paper investigates how the Space Time Adaptive Processing (STAP) radar algorithm and the stripmap Synthetic Aperture Radar (SAR) algorithm can be mapped on a multi-cluster processing system with a reconfigurable optical interconnection system realized by a Micro-Optical-Electrical Mechanical System (MOEMS) crossbars. The paper describes the reconfigurable platform, the two algorithms and how they individually can be mapped on the targeted multiprocessor system. It is also described how these two applications can be mapped simultaneously on the optical reconfigurable platform. Implications and requirements on communication bandwidth and processor performance in different critical points of the two applications are presented. The results of the analysis show that an implementation is feasible with today's MOEMS technology, and that the two applications can be successfully run in a time-sharing scheme, both at the processing side and at the access for interconnection bandwidth.