MAP channel estimation with turbo processing for orthogonal space-time block coded MISO systems

In this paper, we investigate fast time-varying channel estimation for two orthogonal space–time block-coded systems with either two or three transmit antennas and with one receive antenna. We propose an iterative expectation–maximisation-based maximum a posteriori (MAP) semi-blind algorithm for the estimation of the discrete propagation channels seen from the transmit antennas. This algorithm requires a convenient representation of the channel using a discrete version of the Karhunen-Loeve expansion theorem. It incorporates the coded structure of each block in a turbo-processing fashion and uses the MAP training-based channel estimation algorithm for its optimum initialisation. The transmit antennas are assumed to be spatially correlated for the system with two transmit antennas and spatially uncorrelated for the three-antenna system. The validity of the proposed algorithm is highlighted by simulation results. The performance of the proposed receivers, based on MAP channel estimation, for the two systems approaches that of the perfect channel state information receiver, even for very time selective fading channels. Copyright © 2011 John Wiley & Sons, Ltd.