The transition-state pair method for test sequence generation

In this paper, a new approach, called the Transition-State Pair (TSP) method, is proposed for test sequence generation. This method has been implemented using C. The major difference between the TSP method and the W/Wp method is that the TSP method derives a different set for each state as a state´s identity while the W method derives a single set, called W set, for every state and the Wp method derives a different subset of the W set for each state. The TSP method has three advantages over the W method and the Wp method. First, the TSP method can derive a test sequence not only for any minimal and fully specified protocols but also for minimal and partial specified protocols. Second, the TSP method uses less time to derive test sequences. Third, it derives shorter test sequences