Performance of multisensor adaptive MLSE in fading channels

We examine the problem of signal detection over time-varying channels using a multisensor adaptive MLSE detection scheme. We develop analytical expressions for the pairwise probability of error for sequence detection in various channel scenarios. Using the pairwise error probability (PEP) expression for perfect channel tracking we show that there is no error flooring. We then derive the PEP when there are channel estimation errors, and through this analysis we examine the error flooring phenomenon. We derive the PEP for colored Gaussian noise channels when we have a mismatched decoder. Using these results we gain some intuition about the role of channel time-variation, channel estimation errors and noise covariance on the performance of multisensor adaptive MLSE. We also propose an adaptive MLSE scheme which could be used in interference limited environments. We present numerical simulation results to demonstrate the performance of these schemes.