ABSTRACT

Online testing is widely considered to be the most appropriate technique for model-based testing (MBT) of embedded systems where the implementation under test (IUT) is modeled using nondeterministic models [39, 38]. Nondeterminism stems partially from the internal parallel processes of the IUT, timing, and hardware-related asynchronous processes. Other sources of nondeterminism are the higher abstraction level of the model compared to IUT implementation and the ambiguities in the specifications of the IUT. We use the term on the fly in this context to describe the test generation and execution algorithms that compute and send successive stimuli to IUT incrementally at runtime, directed by the test purpose and the observed outputs of the IUT.