Software testing is a very important phase of the software development cycle which ensures that the system developed is reliable and acceptable. Optimizing the test suite size of software eliminates the unnecessary cost and resources that are involved in testing. Sometimes it is not possible to exhaustively test any system due to huge number of test cases. In order to test any system and make it acceptable, combinatorial software interaction testing has been used in several fields. Investigations have concluded that most of the software faults could be identified by pairwise combinatorial interaction testing. Researchers have applied parallel algorithms to various combinatorial optimisation problems and have succeeded in significant time reduction in solving the problems. Large and/or computationally expensive optimization problems sometimes require parallel or high-performance computing systems to achieve reasonable running times.
In this paper we propose a new strategy, “A Parallel Tree Based Strategy for Pairwise Combinatorial Interaction Testing”. The proposed strategy is based on two algorithms, a parallel tree generation algorithm and a parallel cost calculation algorithm, which are used in constructing a test suite with minimum number of test cases. The correctness of the strategy has been proved, and imperial results show that our strategy is efficient in test size reduction.