Using Centroidal Voronoi Tessellations to Improve Adaptive Random Testing
Although random testing is low cost and straightforward, its effectiveness is not satisfactory. To increase random testing effectiveness, researchers have developed Adaptive Random Testing (ART) methods attempting to maximize the test case coverage of the input domain. This project proposes the use of Centroidal Voronoi Tessellations (CVT) to address this problem. CVT can enhance the previous ART methods to improve their coverage of the input space. The generated test cases by other methods act as the input to the CVT algorithm and the output is an improved set of test cases. Therefore, CVT is not an independent method and is considered as an add-on to the previous methods. An extensive simulation study has been performed to demonstrate the effectiveness of CVT against the ART methods. Results from the simulation demonstrate that CVT outperforms previous methods with respect to all studied failure rates and failure patterns. In addition, an empirical runtime comparison between CVT and the other methods has been performed indicating CVT is computationally feasible. To further analyze the CVT method, a randomness analysis was undertaken; this demonstrated that CVT has the same characteristics as ART methods in this regard.