Modeling fault coverage and reliability in a fault-tolerant network

Abstract

A modeling approach incorporating both high-level and lower-level models to evaluate the effectiveness of a reconfiguration scheme used in a fault-tolerant network, is discussed. The high-level model, a small Markov chain, describes the occurrence of link failures and repair, and includes a coverage factor. The coverage factor estimates the probability that the network can successfully reconfigure from complete link failure and depends on the state of the network reconfiguration subsystem when the link failure occurs. The lower-level model describes the state of the network reconfiguration subsystem. After collecting reliability information for the subsystems, the lower-level model can be solved to obtain an estimate of the coverage factor. Once the coverage factor from the lower-level model is inserted into the high-level model, the model can be used to predict the amount of network outage caused by link failures and to help identify the subsystems that are good candidates for upgrade. The recognition process and reconfiguration subsystems of a network are described. >