Minimum Pairwise Coverage Using Constraint Programming Techniques

This paper presented the global constraint pairwise that can be used to enforce the presence of a given pair within a set of test cases or configurations. It also introduced several optimizations for implementing a method that computes the minimum set of test cases that covers pairwise. In addition, the method, seen as a constraint optimization problem, provides a way to compromise between time and efficiency by allowing anytime interruption or time-contract execution. Our approach has been implemented and evaluated on several instances of a test configurations generation problems [5] where input variables are boolean only. We envision to address other instances of these problem where the variables take their values in larger finite domains.