Distributed intelligent call admission control

In this paper, we propose QoS-based distributed intelligent call admission control (DICAC) for wideband multi-service CDMA system. Neural network is employed to predict next-step existing-call interference, fuzzy logic is used to estimate equivalent interference for a new call and make admission decision. QoS performances of all service types in home and neighboring cells are measured as feedback of the system. We introduce distributing idea into intelligent call admission control by defining the concept of equivalent dropping probability of neighboring cells. Simulation results indicate that the proposed algorithm can guarantee the QoS requirements of users, dropping probability and call blocking probability, in both home and neighboring cells