Using the Hopfield model with mean field annealing to solve the routing problem in packet-switched communication networks

Abstract

The performance of the Hopfield neural network with mean field annealing for finding optimal or near-optimal solutions to the routing problem in communication network is investigated. The proposed neural network uses mean field annealing to eliminate the constraint terms in the energy function. Unlike other systems which use penalty constraint terms there is no need to tune constraint parameters and the neural network should avoid the problems of scaling. It also avoids the need to pre-determine the minimum number of hops corresponding to the optimal route. We have obtained very encouraging simulation results for the nine node grid network and fourteen node NFSNET-backbone network.