A Research On The Pairwise Test Data Generation Algorithm Of Parallel Test Based On Solution Space Tree

Combinatorial test,a black box sample technology,can effectively simplify test suite and solve the problem of combinatorial explosion with an idea of combination.Experience shows that pairwise test is able to cover 50%-97% software bugs with high practical values.In addition,the parameters constraints exist commonly with a big reality-oriented meaning to pairwise test case generation.Under the background this paper implements a pairwise test case generation algorithm based on the solution space tree which is aimed at the 2-value constraints and pairwise test case generation.On this basis,the article further proposes a parallel test algorithm which can cut down the test time effectively.First of all,studying the 2-value constraints and their effects on test suite scale;second,the paper maps the SUT to a solution space tree model which is later pruned with the2-value constraints;then traversals the pruned tree model by the backtracking algorithm to generate a small test suite the pairwise coverage criterion;after that,the paper employs a greedy algorithm as a supplement to cover the whole 2-ways combinations;and last,realizes a parallel test algorithm with the parallel between the generation process and the execution and detection process.The simulation results show that: 1)Constraints can reduced the test suite scale to the test case generation algorithm PSST_C that we put forward;2)Under the same conditions,the PSST_C algorithm can produce smaller and more effective test suite compared to ACTS;3)Under the same conditions,the parallel algorithm can effectively shorten the test process time.