Feasible Basic Path Generation Based on Genetic Algorithm

Abstract

In this paper, a feasible basic path generation method based on a genetic algorithm is proposed, which combines a probabilistic statistical method with a genetic algorithm to generate a feasible basic path. First, conditional probability relations and a maximum likelihood estimation method are used to measure the correlation among the conditional statements and determine the mutually exclusive edges in the Decision-to-Decision Graph of the program under test. Then, the feasibility of generating a path is judged by the exclusive edge relation. Finally, a genetic algorithm is used to generate the feasible basic path set of the program under test. The experimental results based on six benchmark programs and three industrial programs show that, compared with the traditional method, this method can effectively improve the coverage rate of feasible basic path generation and reduce the time cost of software testing.