Performance of an array receiver with a Kalman channel predictor for fast Rayleigh flat fading environments

We develop an approach for using an antenna array for tracking fast Rayleigh flat fading channels and suppressing cochannel interference. The fast flat fading process is assumed to be a general autoregressive (AR) process in order to characterize temporal variation of channels and evaluate its effect on the receiver structure and performance. The optimal array receiver structure that minimizes the probability of error for BPSK signals is derived, which includes a Kalman filter to predict the fading channels. A simple integral expression for the probability of error is also derived for the optimal receiver. In particular, we analyze the case with identical shaping filters. An irreducible probability of error is shown to exist due to the prediction error of multiple channels. Another interesting observation from the study is that the diversity gain with m antenna elements in the presence of k interferences is usually greater than (m-k), even in the presence of channel prediction error. Simulations are carried out to verify the theoretical analysis.