Co-Evolutionary High-Level Test Synthesis

The main approach of this paper is utilizing bio-inspired evolutionary strategies for high level test synthesis. In this paper a genetic algorithm (GA) is implemented to schedule a data-flow graph considering latency. Also, module binding is performed with another GA concurrently, considering resource constraints. The register allocation is performed using another GA which minimizes the number of registers. Then a co-evolutionary strategy merges the results of these three solutions, targeting testability improvement. Experimental results show using the proposed approach results in improvement in fault coverage with no or negligible overhead in area and delay.