On antenna selection with maximum ratio transmission

In this paper, we investigate the capacity performance of multiple-input, multiple-output (MIMO) systems with antenna selection. At the transmitter, we assume a single data stream transmission using optimum transmit beamforming, also known as maximum ratio transmission (MRT). The scope of this paper is two-fold: first, to analyze the performance of MRT with antenna selection, and second, to present simple, computationally efficient algorithms that perform near-optimal antenna selection. On the analytical side, we obtain bounds on the ergodicity capacity for MRT with optimum antenna selection over a flat-fading channel. Through simulation, we compare the performance of the MRT with antenna selection and full-feedback systems (where the transmitter knows the channel matrix) and show that having just a few additional antennas enables MRT with antenna selection to perform as well as full-feedback systems for transmission at reasonable signal-to-noise ratios. Next, we propose two new algorithms for performing the antenna selection. These algorithms are significantly more computationally efficient, and have negligible loss of capacity, compared to optimal antenna selection. Through Monte-Carlo simulations, we verify both our analytical results and the performance of the algorithms.