A Greedy Particle Swarm Optimization Strategy for T-way Software Testing

Combinatorial strategies are used as methods or mechanisms for selecting test cases using combinations of test input parameters. We normally want that all t-way combinations of parameter values occur in the test suit at least once. Artificial intelligence base search algorithms have been used within strategies for constructing near optimal test suites. In this paper, we propose a new test generation strategy, for combinatorial testing based on greedy Particle Swarm Optimization. The basic design concepts of the strategy are demonstrated through the paper. The experimental results and comparisons of our strategy showed impressive results as far as the test suite size is considered.