Abstract:
Methods are considered for estimating and increasing reliability of multi-computer system structures with different functioning set-ups. Probabilistic characteristics are determined in simultaneous use of (1) periodic system checking; (2) use of operating time redundancy; and (3) introduction of parallel problem solution with staged comparison of the results. Some typical models of restorable multi-computer systems are investigated. For each model accurate and approximate formulae are obtained of performing specified jobs within specified time of mathematical expectation and time variance in performing jobs of specified size.
Fulltext:
PDF file (1193 kB)