Sequence Generation of Test Case Using Pairwise Approach Methodology

Abstract

There are various instruction specifications in the software system that interacts with each other. While designing software system, software engineers have to face a tedious task of ensuring the validation of every available software use case. If the instruction specifications are huge, the required number of use cases to validate becomes unmanageable within reasonable time budget and therefore designers have to choose one of the options of either delaying the projects or delivering without validating all available test case scenarios. Even for moderately sized software systems, comprehensive testing sometimes becomes impossible due to limited time availability for validation phase. Empirical studies have indicated that most of the faults in software systems often precipitate by interaction between smaller numbers of instruction specifications. If designers can cover all available interactions between all pairs of instructions specifications (instead of comprehensive testing), it can give reasonable confidence to designers about the software attribute within the limited time. This paper presents one such approach which uses the sequence origination approach for pairwise test case origination. This approach makes certain to propagate the required intent of trial run cases which cover all available interactions between all instructions pairs at least once. Trial run selection specification is this approach based on combinatorial testing. The paper also discusses the results with this new approach on one of the candidate software system to demonstrate its efficacy over already existing validation approaches.