GABES: a Genetic Algorithm Based Environment for SEU Testing in SRAM-FPGAs

Testing of FPGAs is gaining more and more interest because of the application of FPGA devices in many safety-critical systems. We propose GABES, a tool for the generation of test patterns for application-dependent testing of SEUs in SRAM-FPGAs, based on a genetic algorithm. Test patterns are generated and selected by the algorithm according to their fault coverage: Faults are injected in a simulated model of the circuit, the model is executed for each test pattern and the respective fault coverage is computed. We focus on SEUs in configuration bits affecting logic resources of the FPGA. This makes our fault model much more accurate than the classical stuck-at model. Results from the application of the tool to some circuits from the ITC'99 benchmarks are reported. These results suggest that this approach may be effective in the inspection of safety-critical components of control systems implemented on FPGAs