Tree Based Test Case Generation and Cost Calculation Strategy for Uniform Parametric Pairwise Testing

Abstract

Problem statement: Although it is very important to test any system extensively it is usually too expensive to do so owing to the cost and the resources that are involved in it. Software testing is a very important phase of software development to ensure that the developed system is reliable. Some systematic approach for testing is essential to test any system and make it acceptable. Combinatorial software interaction testing is one which tests all possible software interactions. This interaction could be at various levels such as two way interaction (pairwise) or three or four or five or six way interactions. Combinatorial interaction testing had been used in several fields. It was reported in literature that pairwise combinatorial interaction testing had identified most of the software faults. Approach: In this study we proposed a new strategy for test suite generation, a tree generation strategy for pairwise combinatorial software testing, with parameters of equal values. The algorithm considered one parameter at a time systematically to generate the tree until all the parameters were considered. This strategy used a cost calculation technique iteratively for each of the leaf nodes to generate the test suite until all the combinations were covered. Results: The experimental data showed that we had achieved about 88% (or more in some cases) of reduction in the number of test cases needed for a complete pairwise combinatorial software interaction testing. Conclusion: Thus, the strategy proposed had achieved a significant reduction in minimizing the number of test cases that was generated.