In this paper we focus on the spatial equalizer design for MIMO systems with Rayleigh fading channels where channel information is not available at the receiver. Training based algorithms have been widely used for such communications scenarios due to its performance superiority compared to blind methods. We investigate how to further improve the performance of the training algorithms by incorporating the statistical information from the unknown sequence. By constructing an ICA-based semi-blind criterion, a steepest decent method is applied to find the desired solution adaptively. To further improve the performance and convergence speed of the proposed method, we propose a technique to find the optimal choice of step size. Furthermore a constrained CRB is developed for the addressed semi-blind signal model to evaluate the performance of the proposed algorithm. Simulation results demonstrate the performance of the proposed algorithms and comparable schemes.
展开▼
