On the robustness of an iterative maximum a posteriori channel estimation for Alamouti's space-time transmit diversity

We evaluate the Alamouti's technique of diversity with a realistic estimation algorithm considering a very fast time-varying channel. We consider the maximum a posteriori iterative semi-blind channel estimation algorithm proposed in Kammoun and Siala (July-August 2003). We show in this paper that it can profit from an optimum initialization by means of the MAP training-based block-by-block channel estimation algorithm. The performance of the channel estimation algorithm with optimum initialization is evaluated by simulation results. We focus also on its robustness to errors in channel and noise statistics.