Pattern generation for Mutation Analysis using Genetic Algorithms

Mutation Analysis (MA) is a fault-based simulation technique that is used to measure the quality of testbenches for mutant detections where mutants are simple syntactical changes in the designs. A mutant is said living if its error effect cannot be observed at the primary outputs. Previous works mainly focused on the cost reduction in the process of MA, because the MA is a computation intensive process in the commercial tool. For the living mutants, to the best of our knowledge, the commercial tool has not addressed the pattern generation issue yet. Thus, this paper presents a Genetic Algorithm to generate patterns for detecting living mutants such that the quality of the verification environment is improved. The experimental results show that more living mutants can be detected after adding the generated patterns in the testbench.