Implementations of Hopfield neural network in communication networks

Hopfield neural network, as a recurrent neural network, shows very good results in solving a lot of different complex computational problems. Starting from its previous modification, and making the new ones, this network was implemented in different types of communication networks in order to solve some real problems. In this paper, the possibility of intelligent decision making of Hopfield neural network, through the three independent implementations, will be presented. Starting from the problem of dynamic routing, one possible solution for multicast routing in telecommunication networks as well as routing in all optical networks will be presented. Beside this, the modifications of Hopfield neural network could be used for planning and projection of the new network infrastructures. One solution for route selection problem in real physical environment will be shown.