Antenna selection and performance analysis of MIMO spatial multiplexing systems

Multiple-input multiple-output spatial multiplexing (MIMO-SM) systems offer an essential benefit referred to as spatial multiplexing gain. Two important signal reception techniques for MIMO-SM systems are the zero-forcing (ZF) and ordered successive interference cancellation (OSIC) as, for example, in the case of the decision-feedback detector (DFD). This thesis studies the communication and signal processing aspects of MIMO-SM. We first investigate the bit error rate (BER) performance of the ZF receiver over transmit correlated Ricean flat-fading channels. In particular, for a MIMO channel with M transmit and N receive antennas, we derive an approximation for the average BER of each sub-stream. A closed-form expression for the optimal transmit correlation coefficient, which achieves the maximum capacity (i.e., uncorrelated case) of two-input two-output spatial multiplexing (TITO-SM) systems, is presented.;We further propose an antenna selection (AS) approach for the DFD over independent Rayleigh flat-fading channels. The selected transmit antennas are those that maximize both the post-processing signal-to-noise ratio (SNR) at the receiver end, and the system capacity. An upper bound on the outage probability for the AS approach is derived. It is shown that the AS approach achieves a performance comparable to optimal capacity-based selection based on exhaustive search, but at a lower complexity.;Finally, we investigate a cross-layer transmit AS approach for the DFD over spatially correlated Ricean flat-fading channels. The selected transmit antennas are those that maximize the link layer throughput of correlated MIMO channels. A closed-form expression for the system throughput with perfect channel estimation is first derived. We further analyze the system performance with pilot-aided channel estimation. In that, we derive a closed-form expression for the post-detection signal-to-noise-plus-interference ratio (SNIR) of each transmitted substream, conditioned on the estimated channels. The derived SNIR is then used to evaluate the overall system throughput. It is observed that the cross-layer AS approach always assigns the transmission to the antenna combination which sees better channel conditions, resulting in a substantial improvement over the optimal capacity-based AS approach. Considering a training-based channel estimation technique, we compare the performance of the proposed cross-layer AS with that of optimal capacity-based AS when employed with a training-based channel estimation. Our results show that the latter is more robust to imperfect channel estimation. However, in all cases, the cross-layer AS delivers higher throughput gains than the capacity-based AS.