Performance modeling of partially self-checking systems

Evaluates the parameters involved in the detection of failures in systems composed of a self-checking part and of non self-checking parts. The self-checking part is supposed to be characterized by a latency detection (eventually equal to zero if there is always an immediate detection). The failure of the other parts are tested by periodically executed test programs. The parameters are the global detection time and the measure of the overhead due to the running of test programs. Optimal values are given as a function of the objectives defined for the system. These results are useful for a wide range of systems, from microprocessor to boards and large systems testing.