Adaptive power control for CDMA systems using linear quadratic Gaussian control

In this paper, we present a robust decentralized method for jointly performing channel estimation and closed loop power control for the reverse link of CDMA networks. Our method is based on linear quadratic control system theory and Kalman filtering and does not require any training symbols for channel or signal to interference ratio (SIR) estimation. The main interest of this new scheme is that it improves the performance of current SIR based power control techniques while avoiding the problem of power escalation, which is often observed in current systems. In this paper, we also provide an analysis on the robustness of this method towards errors on the control command and imperfect call admission control.