A Brief Survey of Fault Tolerant Techniques for Field Programmable Gate Arrays

The main objective of any fault tolerance method is to provide a reliable system. It is important to ensure field programmable gate arrays (FPGAs) are as reliable as possible, especially with new frontiers in application for reconfigurable hardware such as FPGAs. The following survey provides an overview and analysis of standard FPGA fault tolerance techniques: detection, recovery, and mitigation. This survey seeks to provide a foundation for future research and development, such as FPGA use in critical applications at high radiation environments. The following survey provides an analysis on various methods and determines how robust and reliable these methods are through the amount of resources used, speed of detection and accuracy of detection. Overall, for fault detection methods, roving test methods provided the greatest accuracy. For fault repair and recovery, the configuration level repair methods discussed provide the most diverse approach that exploit the reconfigurability feature of FPGAs and have minimal overhead by leveraging available resources. Mitigation, detection, and repair methods operate at different levels and work together to ensure a robust and reliable system.