The semi-blind multiuser receiver based on antenna array

Two semi-blind multi-user receiver´s algorithms based on antenna array are proposed. Here we suppose multiple receiver antennas are our targets. Firstly we estimate the known users´ data and subtract them form the received signal. Secondly we obtain an orthogonal projection matrix (M) of the unknown users´ signal through subspace tracking. Finally by using the matrix M, we transform the detection problem of known users´ data into a minimization problem of quadratic convex function and so that we can solve it with conjugate gradient (CG) algorithm. The initial estimation of known users´ data also serves as the initial value of CG to speed up convergence. We have simulated the performance of the receivers´ algorithm under various conditions and concluded that they are better than other corresponding algorithms.