Action Refinement in Conformance Testing

In model based testing test cases are derived from a model (the specification) of the system we want to test. In general the model is more abstract than the implementation. This may result in test cases that are not executable, because their actions are too abstract; the implementation does not understand them. The standard approach is to rewrite the model by hand to the required level of detail and regenerate the test cases. This is error-prone and time consuming. In this paper we present an approach to automatically obtain test cases at the required level of detail by means of action refinement. Action refinement is a way to add information to the abstract model. It relates actions from the abstract model to concrete actions of the system under test. We apply this approach to a simple case of action refinement, so-called atomic linear input-inputs refinement. In order to reason about correctness between an abstract model and a concrete implementation we introduce a new implementation relation. We show that this relation is equivalent with the uioco implementation relation on the refined model. Furthermore we show under which conditions the refinement of a complete abstract test suite is again complete.