Self-testing of multiprocessor systems with regular diagnostic connections

Abstract

We propose an approach to organizing self-testing in a multiprocessor system under certain constraints; in particular, we use special diagnostic graphs, and the number of failures does not exceed a certain value T. We show that a test experiment where each of n processors is tested by two others lets one determine the state of all except possibly two processors for T = 4 and except one for T = 3. The total number of tests does not exceed 2n + 2.