Blind channel equalization based on second order statistics

In this paper, we present new approaches to blind channel estimation and equalization based on second order statistics (SOS). We first consider the case of minimum phase channels where the equalizer is designed based on the criterion of autocorrelation matching. We cast the problem as a convex optimization program that can be efficiently solved using interior point methods. Then we consider the equalization of single-input multiple-output (SIMO) channels. Due to oversampling, the equivalent channel matrix possesses a particular structure which enables us to estimate the channel based only on the information contained in the covariance matrix at zero delay. Simulation examples are provided to demonstrate the performance advantage of the proposed algorithms compared to existing techniques.