Diagnosing Multiple Faults in Finite State Machines

In this paper, we propose a generalized diagnostic algorithm for the case where more than one fault (output and/or transfer) may be present in the transitions of a system represented by a deterministic finite state machine (FSM). If existing faults are detected, this algorithm permits the generation of a minimal set of diagnoses, each of which is formed by a set of transitions (with specific types of faults) suspected of being faulty. The occurrence in an implementation, of all the faults of a given diagnosis, allows the explanation of all observed implementation outputs. The algorithm guarantees the correct diagnosis of certain configurations of faults (output and/or transfer) in an implementation, which are characterized by a certain type of independence of the different faults. We also propose two approaches for selecting additional test cases, which allows the reduction of the number of possible diagnoses. A simple example is used to demonstrate the different steps of the algorithm.