Chaotic neural network method to control ATM traffic

A chaotic neural network method is set forth to model and predict videoconference VBR traffic. The chaotic theory is employed to analyze the data and a neural network is used to model and predict the traffic. Based on the prediction of the VBR traffic, we estimate the available bandwidth and feed back explicit rate (ER) to the ABR source to control ABR traffic with VBR traffic background. The simulation results showed that the performance of the network is greatly improved with ER estimation via the chaotic neural network method.