Improving input-output conformance testing theories
N Neda Noroozi · 2014
Software quality has become an increasingly important concern in software development, as a consequence of the ever-ascending trend of using software system in critical systems. Testing as a technique for establishing a certain level of software quality has received much attention in the past decades. Model-based testing is a structured approach to testing. Using model-based testing the process of generating test-cases and predicting the correct outcome of test-cases can be mechanized. By using rigorous models for system behavior, model-based testing is formalized in terms of a mathematical notion of conformance. Input-output conformance (ioco) is a wide ly-s tud ied and commonly-used conformance relation. The ioco relation has some limitation in testing concurrent systems. Such systems are composed of interacting components which often communicate asynchronously with their environment. However, the ioco testing theory was developed based on the assumption that a tester can always communicate with an implementation under test synchronously. It is also well-known that the ioco relation does not have the compositionality property. In this thesis theoretical foundations are developed to provide solutions to make the ioco relation suited for testing concurrent systems.