Characterizing Faults on Real-Time Systems Based on Grid Automata

Abstract

Real-time systems are, in general, critical systems that interact with the environment through input and output events regulated by time constraints. The testing activity on systems of this nature requires rigorous approaches due to their critical aspects. Model-based testing approaches rely on formalisms that provide more reliability to testing activities. However, a model-based testing approach for real-time systems depends on techniques that can deal with continuous evolution of time appropriately. Several testing approaches apply discretization techniques in order to represent continuous behavior of timed models. Test suites can then be extracted from discretized models to support conformance testing between specifications and their respective implementations. Therefore an evaluation of test suites considering a fault coverage is an important task, but rarely addressed by model-based testing approaches for real-time systems. In this work we propose a systematic strategy to identify faults in TIOA models based on their corresponding discretized models. We precisely define a fault model to support model-based testing activities such as coverage analysis and test case generation.