A NONLINEAR FILTERING APPROACH TO SYMBOL DETECTION AND STATE ESTIMATION IN RICE/RAYLEIGH CHANNELS

摘要

In this paper we propose a stochastic nonlinear filtering approach to the problems of symbol detection and multiplicative distortion estimation of a digitally phase-modulated carrier in Rice/Rayleigh frequency flat-fading channels with additive white Gaussian noise. The derived receiver is an open-loop structure consisting of a bank of stochastic nonlinear filters and a decision algorithm based on the associated innovations processes. This receiver performs symbol-by-symbol detection while keeping track of the carrier random amplitude and phase within the symbol interval. Good results in terms of error probabilities are obtained even in adverse situations, corresponding to fast-fading and/or significant carrier phase drifts.