A Reliable Reconfiguration Controller for Fault-Tolerant Embedded Systems on Multi-FPGA Platforms

This paper proposes the design of a controller managing the fault tolerance of multi-FPGA platforms, contributing to the creation of a reliable system featuring high flexibility and resource availability. A fault management strategy that exploits the devices' reconfiguration capabilities is proposed, the Reconfiguration Controller, focus of this paper, is the main component in charge of implementing such strategy. The innovative points raised by this work are 1) the identification of a distributed control architecture, allowing the avoidance of single points of failure, 2) the management of both recoverable and non-recoverable faults, and 3) the definition of an overall reliable multi-FPGA system.