Towards Scalable Robustness Testing

Several approaches have been developed to assess the robustness of a system. We propose a model-based approach to scalable testing the robustness of a software system using event sequence graphs (ESG) and decision tables (DT). Elementary modification operators are introduced to manipulate ESGs and DTs resulting in faulty models. Test cases generated from these faulty models are applied to the system under consideration to check its robustness. Thus, the approach enables the quantification of robustness with respect to a universe of erroneous inputs.