Time varying vector channel estimation for adaptive spatial equalization

New blind adaptive spatial equalization techniques are proposed for severe multipath fading with co-channel interference. In rapidly fading environments, equalizer performance suffers with existing algorithms because the desired and interference channels change significantly over the update time interval. It is shown that for fast fading channels, conventional squared error equalizer cost functions can yield poor results. A new algorithm is proposed based on a variation of the SINR cost function coupled with optimal time varying channel estimation to improve signal tracking. In simulations, the algorithm achieves theoretically optimum time varying equalizer performance in extreme interference channels with narrow band FM signals and high velocity portables. These results are contrary to the widespread belief that high performance blind adaptive array processing is not possible for AMPS with severe combinations of Doppler spread, angle spread, and interference power.