Equalization of Time-Variant Communications Channels via Adaptive AR Prefiltering

For the purpose of time-variant channel equalization, a novel structure of equalizer is proposed. An autoregressive (AR) filter is used as the prefilter and a finite impulse response (FIR) filter is cascaded. The AR prefilter is constructed from channel estimation results adaptively, and the cascaded FIR filter behaves as an adaptive FIR equalizer. It is shown that the resulting infinite impulse response (IIR) equalizer is closely related with the IIR-type Wiener filter. Simulations demonstrate that the proposed equalizer provides better performance than the FIR equalizer with the least mean square adaptation in time-variant environments.