A combinatorial intelligent computing approach for resolving the delay and flow assignment in computer communication networks

In this paper, the average end-to-end delay of packets in a computer communication network is investigated. The network delay depends mainly on the transmission route and its flow assignment, and can be decreased through a suitable route selection and flow assignment scheme. The route selection is a static virtual-circuit assignment issue and can be consider as a 0–1 quadratic programming problem with multiple constraints. To this end, we first employ the genetic algorithm to initialize the route, then employ a new artificial neural network model to solve the delay problem, resulting in further reduction in total network delay. The simulated results confirm that this combinatorial neural intelligent computing approach is feasible and effective.