Skip to Main Content
A search method based on particle swarm optimisation to search simultaneously the energy function coefficients and the parameters of the Hopfield neural network (HNN) algorithm is proposed. Experiments using three different communication networks are presented. The results show that the proposed approach achieves better results than existing algorithms that employ the HNN for routing. Some results show that the algorithm tends to be quicker than Dijkstra for larger communication networks.
Date of Publication: July 31 2008