Search-Based test case prioritization for simulation-Based testing of cyber-Physical system product lines

Abstract

Cyber-Physical Systems (CPSs) integrate computation with physical processes. These systems are usually highly configurable to address different customer needs and are evolving to be CPS product lines. The variability of CPS product lines is large, which implies that they can be set into millions of configurations. As a result, different cost-effective methods are needed to optimize the test process of these systems. We propose a search-based approach that aims to cost-effectively optimize the test process of CPS product lines by prioritizing the test cases that are executed in specific products at different test levels. The prioritized test suite aims at reducing the fault detection time, the simulation time and the time required to cover functional and non-functional requirements.We compared our approach by integrating five search algorithms as well as Random Search (RS) using four case studies. As compared with RS, the search algorithms managed to reduce fault detection time by 47%, the simulation time by 23%, the functional requirements covering time by 22% and the non-functional requirements covering time by 47%. Moreover, we observed that the performance of search algorithms varied for different case studies but the local search algorithms were more effective than the global search algorithms.