Dynamic routing in ATM networks with effective bandwidth estimation by neural networks

In this paper, we propose a dynamic routing method for VP-based ATM networks. We use a modified least loaded routing algorithm which employs neural networks(NNs) to estimate the required effective bandwidths of traffic sources. A large set of training data that represents different traffic patterns has been used to ensure that the NN can generalize and produce accurate results. Simulations show that the neural estimation is more accurate and hence leads to much lower call blocking probabilities than the conventional routing method.