FPGA based distributed self healing architecture for reusable systems

摘要

Creating an environment of "no doubt" for com-puting systems is critical for supporting next generationscience, engineering, and commercial applications. Withreconfigurable devices such as Field Programmable GateArrays (FPGAs), designers are provided with a seductivetool to use as a basis for sophisticated but highly reli-able platforms. Reconfigurable computing platforms poten-tially offer the enhancement of reliability and recovery fromcatastrophic failures through partial and dynamic reconfig-urations; and eliminate the need for redundant hardware re-sources typically used by existing fault-tolerant systems. Wepropose a two-level self-healing methodology to offer 100%availability for mission critical systems with comparativelyless hardware overhead and performance degradation. Ourproposed system first undertakes healing at the node-level.Failing to rectify the system at the node-level, network-level healing is then undertaken. We have designed a sys-tem based on Xilinx Virtex-5 FPGAs and Cirronet wirelessmesh nodes to demonstrate autonomous wireless healing ca-pability among networked node devices. Our prototype is aproof-of-concept work which demonstrates the feasibility ofusing FPGAs to provide maximum computational availabil-ity in a critical self-healing distributed architecture.