DETECTION AND IDENTIFICATION OF MALFUNCTIONS IN DISTRIBUTED CONTROL COMPUTER-SYSTEMS WITH PROGRAM-CONTROLLED FAULT-AND-FAILURE TOLERANCE

A method is considered for ensuring the fault-and-failure tolerance of a computer system by way of the simultaneous fulfillment of the same task by a pew computers with subsequent interchange of the results and the selection of a correct result. A method is proposed for the distr ibuted majorization of information with the consistent and synchronous detection and identification of a single malfunction (a fault or fail ure) at the site of its origin. A is admissible that a faulty device c an be either a computer or its transmitter interface tied to the inter computer communication. channel of the bus architecture. A barge detec tion coverage of fault symptoms and a barge coverage of their registra tion during identification are made possible by means of the continuou s, end-to-end functional diagnostication. An example is given.