Recursive Algorithms for Estimating Multiple Co-Channel Digital Signals Received at an Antenna Array

We have recently proposed a new maximum likelihood approach for separating multiple co-channel digital signals received at an antenna array. This approach exploits the-nite alphabet property of digital signals to simultaneously estimate the array response matrix A and symbol matrix S, given the data matrix X = AS. In 1, 2], we presented two eecient block algorithms for computing the estimates: Iterative Least-Squares with Projection (ILSP) and Iterative Least-Squares with Enumeration (ILSE). These algorithms are based on the alternating projections technique whereby the ML criterion is minimized successively, rst with respect to the variable A, and then with respect to S. In this paper, we consider the recursive extensions of these algorithms. We also present a new approach which minimizes the ML criterion directly, assuming that the previous symbols have been estimated correctly. Simulation demonstrate the promising performance of these recursive techniques.