Performance evaluation on hybrid fault diagnosability of regular networks

Diagnosability is an important metric to the capability of fault identification for multiprocessor systems. However, most researches on diagnosability focus on vertex fault. In real circumstances, not only vertex faults take place but also edge malfunctions may arise. Recently, a kind of new diagnosability under hybrid fault circumstances, called h-edge tolerable diagnosability, has been proposed and the h-edge tolerable diagnosability of n-dimensional hypercube under the PMC model and MM* model is determined to be n - h for n >= 4 and 1 <= h <= n - 1. In this work, we propose a general approach to determine the h-edge tolerable diagnosability of general regular networks. We show that the h-edge tolerable diagnosability of a t-regular t-connected network with N processors under the PMC model (resp., MM* model) is t-h for t >= 2 and 1 <= h <= t - 1 if N >= 2 (t-h) +1 (resp., N >= 2(t - h) + 3). Moreover, if G = (V, E) is a t-regular t-diagnosable network under the PMC (resp., MM*) model (where t >= 2), then t(h)(e)(G)= t - h for 1 <= h <= t -1 under the PMC (resp., MM*) model. (C) 2019 Elsevier B.V. All rights reserved.