Cost-oriented component redundancy allocation for a communication system subject to correlated failures and a transmission reliability threshold

Abstract

Transmission quality of a communication/computer system is a high-level objective of system supervisors. Therefore, transmission reliability improvement or optimization is an important issue for many organizations. One way to maximize transmission reliability is to model the system as a stochastic communication network including arcs and nodes and then determine the optimal component redundancy allocation. However, modern components are highly reliable. Thus, a decision maker may be more concerned about cost than reliability. This article considers cost-oriented component allocation subject to a reliability threshold and correlated failures characterized by a correlated binomial distribution model. To solve this problem, we employ a genetic algorithm to search for the optimal component redundancy allocation possessing minimal allocation cost. The computational efficiency of the genetic algorithm–based method is demonstrated through several benchmark networks and compared against several popular soft computing algorithms.