Combinatorial t-way test suite generation using an improved asexual reproduction optimization algorithm

To ensure the correctness and quality of a software system, it is desirable to test all possible combinations of the input parameters under various configurations. But the exhaustive testing of software systems with a large number of input parameters is practically impossible due to the combinatorial explosion problem. In order to address and mitigate this problem, combinatorial t-way strategy can be used to generate an array of test cases covering all combinations of only t input parameters. Since the minimum covering array generation (MCAG) is an NP-hard optimization problem, recently some strategies based on meta-heuristics have been used to solve this problem. Despite the usefulness of these strategies, they cannot completely solve the MCAG problem in systems with a large number of input parameters. This paper solves the mentioned MCAG problem by adapting the Asexual Reproduction Optimization (ARO) algorithm. Additionally, the ARO algorithm is improved (named ImpARO) by changing the population size, the mutation and crossover functions. The statistical analysis of the experimental results on several different configurations shows that ImpARO outperforms ARO, GALP, GS, BAPSO, DPSO, WOA, and GSTG as meta-heuristics, TConfig as a mathematical strategy, and PICT and IPOG as greedy strategies by 89%, 46%, 68%, 65%, 69%, 62%, 65%, 64%, 65%, and 59%, respectively. Moreover, various experiments show that ImpARO has faster convergence speed compared to ARO.