Automated requirements-based test case generation

Abstract

Black-box testing is a technique in which test cases are derived from requirements without regard to the internal structure of the implementation. In current practice, the black-box test cases are derived manually from requirements. Manually deriving test cases from requirements is a costly and time consuming process. In this paper, we present the notion of autogenerating black-box test cases from requirements, that can result in dramatic time and cost savings.To accomplish this, we use requirements formalized as temporal logic properties. We define coverage metrics directly on the structure of the formalized requirements, and use an automated test case generation tool, like the model checker, to generate test cases from formal requirements that satisfy the desired criteria. To evaluate the effectiveness of black-box test suites generated in this manner, we measure the implementation coverage achieved by the test suites, and their fault-finding effectiveness. In [11], we conducted a preliminary investigation using a close to production model of a Flight Guidance System developed at Rockwell Collins Inc. We autogenerated requirements-based test suites for three different requirements coverage criteria and evaluated them by measuring the implementation coverage achieved.