AN ELECTION ALGORITHM FOR A MULTIPROCESSOR CONTROLLED DIGITAL EXCHANGE

Abstract

A multiprocessor controlled distributed digital exchange is studied in case of some failure types. The recovery from a failure is based on temporary suspension of the normal operation and the active processors elect a coordinator controlling the reconfiguration of the system. The algorithm developed for such an election is used at initial (cold) start of the system and in case of reintroducing one or more processors repaired as well. The failure environment is given and the assertions are proved verifying the algorithm described.