Hybrid Particle Swarm and Ranked Firefly Metaheuristic Optimization-Based Software Test Case Minimization

Abstract

Software testing is a valuable and time-consuming activity that aims to improve the software quality. Due to its significance, combinatorial testing focuses on fault identification by the interaction of small amount of input factors. But, deep testing is not sufficient due to time or resources availability. To select the optimal test cases with least computation time, Hybrid Multi Criteria Particle Swarm and Ranked Firefly Metaheuristic Optimization(HMCPW-RFMO) technique are introduced. Initially, the population of the test cases is randomly initialized. Then the fitness is calculated by the pairwise coverage, execution cost, fault detection capability and average execution frequency. RFM approach starts with ‘n’ fireflies. The light intensity of each firefly gets initialized.If the light intensity of one firefly is minor than the other one, it moves near the brighter one. Next, the rank is given to the firefly based on their light intensity. Lastly, the high ranked firefly is chosen as a global best solution.The result reveals that HMCPW-RFMO technique improves the software quality.