Multi-objective network reliability optimization using evolutionary algorithms

This work presents a new multiple objective evolutionary algorithm to solve three well known network reliability allocation problems considering different conflicting objectives to be optimized simultaneously. The new algorithm is applied in the design of a telecommunication network that is formed for several stations or nodes interconnected by telecommunication links or paths. The problem presented in this work involves finding which links to activate in order to obtain connectivity in the nodes. The number of nodes that need to be connected depends of the case that is being evaluated. The three network reliability problems considered are: all-terminal, k-terminal, and two-terminal. Network reliability evaluation represents an NP-hard problem. In literature two approaches have been presented: exact reliability calculation and reliability estimation. Because of the impracticality of exact calculation in networks of moderate to large size Monte Carlo simulation is used to estimate network reliability. The new algorithm presented in this work is based on evolutionary computation and the objective functions considered are the maximization of system reliability, the minimization of system cost and, the minimization of system weight. The solution to this multiple objective problem is a set of Pareto solutions