A Fault-Detection Approach to the Conformance Testing of Nondeterministic Systems

The classical fault-detection approach to conformance testing for deterministic systems is based on state identification by means of diagnosis sequences. This paper extends this approach to set identification for nondeterministic systems in which an action may reach more than one state. It presents a new scheme for testing and deriving diagnosis sequences in terms of both specified and unspecified actions of a nondeterministic labeled transition system. New definitions and derivation methods for generalized unique input/output sequences and distinguishing sequences for set identification are proposed. The methods preserve failure equivalence and require minimization of the set space with respect to set-observational equivalence.