Statistical MIMO antenna sub-set selection with space-time coding

Multiple input multiple output antenna subset selection is a low cost low complexity technique with the benefits of multiple antennas. This paper addresses the problem of statistical MIMO antenna sub-set selection with space-time coding. The antennas are chosen based on second order channel statistics, the goal being to minimize the average probability of error. We show that the optimal antenna set maximizes the determinant of the co-variance of the vectorized channel. The derived selection rule allows simultaneous (joint) selection of optimal transmit and receive antennas. We discuss propagation scenarios in which the selection rule decouples so that joint selection can be carried out by performing transmit selection independent of the receive antennas and vice versa. The selection gain is quantified as the performance improvement due to transmission/reception on the optimal set instead of any other selection. We show the possibility of both coding gain as well as diversity gain. Finally, we support our analysis with simulations.