A power control scheme with link gain prediction using PRNN/ERLS for DS-CDMA cellular mobile systems

Abstract

In this paper, a link gain prediction-based power control scheme is designed for DS-CDMA cellular mobile systems. The link gain prediction can help to remove the interference of the power control adjustment itself. The pipeline recurrent neural network (PRNN) with extended recursive least square (ERLS) is adopted for the prediction. This PRNN/ERLS predictor possesses infinite memory of past signals so that it can capture precisely the signal correlation and improve the delay compensation of power control. Simulation results show that the link gain prediction-based power control scheme using PRNN/ERLS has lower outage probability than the received SIR prediction-based power control scheme using grey prediction method [5] and improves the system capacity by an amount of 41.6%.