| Modern wireless communication technology has been greatly impacted by the advances in two research disciplines. On one hand, the theoretical setups in information theory originated from Shannon's theorem has influenced the development of major communication techniques in the past half century. On the other hand, the every-18-month doubling in hardware capacity per dollar, as predicted by Moore's Law, accelerates the progress in DSP technology. Keeping step with these two fronts, problems that stem from practical applications in wireless communications call for a fusion of the two. The main objective of this dissertation is to help bridge the gap between the theoretical advances in information theory and the implementation realities supportable by the current signal processing technology.; MIMO (Multiple-Input-Multiple-Output) techniques, utilizing multiple antennas at both the base station and the mobiles users, are rapidly becoming one of the most promising solutions to facilitate better cell coverage, higher data rates and more reliable transmission. On one hand, the diversities from time, space, and frequency inherent in multiple antenna systems can be treated in a unified framework to make possible a theoretically optimal data rate. On the other hand, many practical system design issues (including, e.g., "reliability" and "complexity") must be resolved before such an optimal spectral efficiency can be fully realized. In designing a smart antenna system, we must deal with inevitable uncertainties and impairments such as the time-varying environments, co-channel interference, existence of multi-paths, and noise disturbances. This will necessitate novel signal processing solutions to overcome the interference caused by the resource reuse and the distortion of the wireless channels, while consuming minimum power and computational cost.; This dissertation addresses several topics in the exploration of space-time diversity in multiple antenna wireless systems: performance evaluation, algorithm design, and implementation cost. The optimal spectral efficiency can be mathematically established from an information theoretic point of view. We provide a unified analysis on the capacity performance of different MIMO equalizers for both flat and frequency-selective fading channels. (Abstract shortened by UMI.)... |
