| The ultimate goal of next generation wireless communications is to provide universal personal and multimedia communications without regard to mobility or location. Future services including a wide range of applications such as high-quality voice, data, and video, etc, will require transmission rates of several Mbps over a limited frequency spectrum. Due to co-channel interference (CCI) and intersymbol interference (ISI) induced by multipath fading, however, there is an irreducible error floor that imposes a limit on the maximum attainable transmission rate.; In this thesis, we propose solutions using multiple-input multiple-output (MIMO) smart antenna systems for single and multi-user wireless communications. The problems are studied under the assumption that the transmitter knows the channels, and the objective of the thesis is to investigate advanced space-time processing that optimizes the system performance. Three new systems are described here, one single user and two multi-user systems. They are smart base and mobile (SBM) antenna system, maximum transmit signal-to-interference plus noise (MTxSINR) ratio system, and joint multiuser MIMO system (JMMS).; For single user communications we have derived a closed-form solution for the transmit and receive antenna weights, operating jointly, to maximize the average packet SINR under frequency selective fading channels in the presence of CCI. Simulation results demonstrate that we can get lower probability of bit error for a given SNR and SIR in frequency selective fading channels than that in flat fading channels. For multi-user communications, we study the optimization problem of spectral efficiency of a multi-user MIMO system for communication from one base station to many mobile stations. To solve the global optimization of multi-user MIMO antenna weights, we have developed a closed-form solution namely MTxSINR which is based on maximizing the lower bound of the product SINR. Results show that multi-user optimization outperforms significantly single-user optimization. To further improve the performance of a multi-user system, a novel simultaneous channel diagonalization has been studied. The objective of such multi-channel diagonalization is to partition or distribute multi-user signals into disjoint space and the resultant channel gains are maximized to optimize the overall system capacity under the constraint of a fixed transmit power. Simulation results reveal that enhanced performance is possible. |
