Assessing reliability within error rate and time constraint for a stochastic node-imperfect computer network

Abstract

In order to maintain the stability of a computer network, the assessment of system reliability is an important issue for the supervisor. A computer network can be modeled as a network topology with edges and nodes, in which each edge denotes a transmission line such as coaxial cables or fiber cables, and each node denotes a transmission station such as router or switch. For a real-life computer network, the capacity of each component (edge or node) should be stochastic owing to complete failure, partial failure, etc. Hence, the computer network with imperfect components is also stochastic and is named a stochastic node-imperfect computer network. Different from the quickest path problem considering deterministic capacity, this article concentrates on a stochastic node-imperfect computer network to evaluate the probability that d units of data can be transmitted within tolerable error rate and time constraint. Such a probability, called system reliability, is a performance indicator to provide to managers for further improvement. An efficient algorithm in terms of the approach of minimal paths to evaluate the system reliability is proposed.