A dynamic neural estimator for admission control in ATM network

In call admission control of ATM networks, it is difficult for the conventional methods to judge the accepting boundary accurately and dynamically, for the imprecise description of the traffic parameters and the different requirement of the allowed QoS. We propose a neural network structure as an intelligent control scheme to perform ATM admission control. The neural estimator can learn the probability distribution of the CLR, thus can control the ATM traffic very accurately and dynamically. The discrete structure of the neural estimator makes it easy to learn and operate. The trained neural network can also be used as a buffer estimator in the reference design of ATM system. The simulation result shows the advantage of this neural estimator