A diagnosis algorithm by using graph-coloring under the PMC model

Abstract

Fault diagnosis is important to the design and maintenance of large multiprocessor systems. PMC model is the most well known and widely studied model in the system level diagnosis of multiprocessor systems. Under the PMC model, a diagnosis algorithm based on some graph-coloring techniques has been proposed in this paper. Given a syndrome $$\sigma $$ź, the first part of the algorithm can locate all the definitely faulty vertices. Then in the second part of the algorithm a diagnosis graph corresponding to the syndrome can be constructed and the suspicious faulty sets can be determined by finding the maximal independent sets of the diagnosis graph. A weight is assigned to each suspicious faulty vertex set which can measure its occurring probability. The algorithm is shown to be correct, not based on any conjecture and can be applied to the fault identification for any multiprocessor system.