Abstract:
Principles of diagnosing a failure-proof duplicated selfchecking system with sliding redundancy are considered. A computer model is described whose properties ensure detection of the faulty computer in a sequence of operations. The consistency of the sequence is proved.