NEURAL NETWORKS FOR SHORTEST-PATH COMPUTATION AND ROUTING IN COMPUTER-NETWORKS

Recently neural networks have been proposed as new computational tools for solving constrained optimization problems. This paper is concerne d with the application of neural networks to the optimum routing probl em in packet-switched computer networks, where the goal is to minimize the network-wide average time delay. Under appropriate assumptions; t he optimum routing algorithm relies heavily on shortest path computati ons that have to be carried gut in real time. For this purpose an;effi cient neural network shortest path algorithm, that is an improved vers ion of previously suggested Hopfield models, is proposed. The general principles involved in the design of the proposed neural network are d iscussed in detail. The computational power of the proposed neural mod el is demonstrated through computer simulations. One of the main featu res of the proposed model is that it will enable the routing algorithm to be implemented in real time and also to be adaptive to changes in link costs and network topology.