Adaptive power control on the reverse link for CDMA cellular system

Abstract

Power control is of paramount important to combat the near-far problem and co-channel interference in a CDMA cellular system. While the conventional fixed-step power control scheme is difficult to compensate the fast fading channel dynamically, a fuzzy neural network (FNN) based power controller is proposed to smoothen out the fast fading and keep the received signal power from each user constant at the base station. Simulation results show that the FNN power controller achieves better traffic capacity and control performance than a fixed-step controller due to the inherent ability of FNN to identify the fast time-varying characteristic of the inverse channel. Furthermore, the FNN scheme is suitable for on-line system identification and adaptive control by neural network training.