Skip to Main Content
As an activity to ensure quality and conformance, testing is one of the most important activities in any software or hardware product development cycle. Often, the challenge in testing is that the system may support a wide range of configurations. Ideally, it is desirable to test all of these configurations exhaustively. However, exhaustive testing is practically impossible due to time and resource limitations. To address this issue, there is a need for a sampling strategy that can select a subset of inputs as test data from an inherently large search space. Recent findings demonstrate that t-way interaction testing strategies based on artificial intelligence (i.e. where t indicates interaction strength) have been successful to obtain a near optimal solution resulting into smaller test set to be considered. Motivated by such findings, we have developed a new test generation strategy, called Particle Swarm Test Generator (PSTG). In this paper, we discuss the design of PSTG and demonstrate our preliminary test size reduction results against other competing t-way strategies including IPOG, WHITCH, Jenny, TConfig, and TVG.