An adaptive parametric prediction method for mobile MIMO wireless systems

In this paper, we investigate the prediction of mobile MIMO channels with varying multipath parameters. Based on the PAST algorithm, we propose a multidimensional adaptive ESPRIT approach for jointly tracking the evolution of the Doppler frequencies and spatial directions of arrival and departure of the propagation paths. Future states of the channel are predicted using the last estimate of the propagation parameters. We show via simulation that the proposed adaptive method outperforms existing static approaches with varying channel parameters. Our results indicate that the performance improvement from parameter tracking is dependent on the rate of variation of the underlying multipath parameters.