An FPGA-based fail-soft system with adaptive reconfiguration

Increasing autonomous fault-tolerant FPGA-based systems' lifetime

“…Similar limitations can also be noted in the other approach [10]; few details are provided on the strategy for the definition of the self-checking areas and for the floorplanning, which considers a very simplistic device model containing only basic configurable logic blocks.…”

“…In literature, very few works address permanent faults in the design of reliable FPGA-based systems ( [7], [8], [9], [10]). These studies can be classified into two categories, based on the adopted strategy to define the alternative necessary bitstreams: i) on-line bitstream computation, that dynamically creates the recovery configurations, and ii) off-line bitstream computation, that creates the alternative configurations during the design phase.…”

“…The off-line strategy ( [9], [10]) requires the storage of all the bitstreams that might be needed during the system's lifetime. A large memory could be required, based on the number of permanent faults that can be tolerated in relation with the fault localization strategy.…”

“…Among the alternative approaches available in literature, few details are actually provided (e.g., [9], [10]) and no comparison can straightforwardly be carried out. However, we took the strategy presented in [9] and used our flow to generate implementations following their guidelines, to make a comparison.…”