Performance analysis of adaptive DFE-RLS in mobile fading channels

Abstract

An extensive analysis of using the decision feedback equalizer (DFE) that uses a recursive least squares (RLS) adaptation algorithm to equalize multipath fading radio channels in general and digital cellular radio (DCR) channels in particular is performed. Throughout this analysis, the effect of the fade rate, multipath spread, equalizer size, and tracking speed on convergence rate and steady state behavior of a DFE-RLS is examined. Also, for a given fade rate, a compromise between fast adaptation (necessary to track variations in the input statistic) and slow adaptation (necessary to contain noise in the adaptive process) is performed. In addition, for a proposed mobile communication system for North America, recommendations are also documented for some system parameters to ensure an acceptable performance.