System-Level Fault Diagnosis Using Comparison Models: An Artificial-Immune-Systems-Based Approach

Abstract

The design of large dependable multiprocessor systems requires quick and precise mechanisms for detecting the faulty nodes. The problem of system-level fault diagnosis is computationally difficult and no efficient and generic deterministic solutions are known, motivating the use of heuristic algorithms. In this paper, we show how artificial immune systems (AIS) can be used for fault diagnosis in large multiprocessor systems containing several hundred nodes. We consider two models—the simple comparison model and the generalized comparison model (GCM), and we propose AIS-based algorithms for identifying faults in diagnosable systems, based on comparisons among units. We performed experimental analysis of these algorithms by simulating them on randomly generated diagnosable systems of various sizes under various fault scenarios. The simulation results indicate that the AIS-based approach provides an effective solution to the system-level fault diagnosis problem.