Neuro-Fuzzy admission control in cellular networks

In this paper, a methodology is presented for designing an adaptive fuzzy logic controller based on neural networks. The neuro-fuzzy controller is first trained using data from an approximate analytical model of a cellular network then the controller is fine tuned and adapted to the unique cell dwell time and call holding time distributions of a particular cell in the network. Different cell dwell time distributions are considered for training the neuro-fuzzy controller. A neuro-fuzzy method that only relies on a limited amount of measured data for training purposes is also presented.