A Search-Based Automated Test-Data Generation Framework for Safety-Critical Systems

Abstract

This paper presents the results of a three year research program to develop an automated test-data generation framework to support the testing of safety-critical software systems. The generality of the framework comes from the exploitation of domain independent search techniques, allowing new test criteria to be addressed by constructing functions that quantify the suitability of test-data against the test-criteria. The paper presents four applications of the framework - specification falsification testing, structural testing, exception condition testing and worst-case execution time testing. The results of three industrial scale case-studies are also presented to show that the framework offers useful support in the development safety-critical software systems.