It is estimated that the number of internet users and the number of cellular phone users worldwide will soon reach more than one billion. Now, wireless communication is witnessing a rapid growth in technology, markets, and range of services. The trends include requiring wireless communication systems to link with the wireline infrastructure for internet accessing, to improve the robustness in time-variant multipath environment for mobile communication, to reduce the system complexity for economic and power constraints, and to enhance the spectrum efficiency of high-rate transmission over limited bandwidth.; An attractive approach for providing low-complexity and reliable transmission over frequency-selective fading multipath environment is the use of orthogonal frequency division multiplexing (OFDM) modulation. Another unrelated promising approach for improving bandwidth efficiency, transmission speed, and reliability is the use of antenna arrays. The first goal of this thesis is to explore methodologies for integrating these two approaches. Consequently, we propose advanced space-time signal processing algorithms to combine OFDM modulation and multiple-antenna systems. The research begins with investigating appropriate channel and signal models to characterize the time-variant multipath nature of wireless propagation. We then apply various signal processing algorithms to improve the system performance in terms of bit error rate (BER) and transmission rate. Analytical evaluations and simulations have been performed to investigate the system performance. To save computational work and reduce operational time, various approaches are proposed based on QR decomposition and parallel processing architectures. To show the performance of the OFDM multiple antenna system in the real world, a practical example applying OFDM multiple-antenna system to avionics telemetry is given.; To further enhance the channel capacity and bandwidth efficiency of OFDM multiple-antenna systems, optimal power and bit allocation methods subject to power and quality of service constraints are derived. The optimal solution is the 2-D water-filling form embedded in the space and frequency domains. To attain the performance in the time-varying environment, various channel estimators for channel tracking and optimal training sequences for channel acquisition are also designed. We evaluate the system performance in terms of BER, channel estimation error, and outage capacity under different angle spread, number of antennas, and Doppler spread conditions.; Space time signal processing can be applied not only to wireless communications but also to acoustics and seismic problems. Therefore, the second part of this dissertation applies space time signal processing algorithms with acoustic or seismic sensor-array to perform sources localization, tracking, separation, extraction, enhancement, classification and recognition. The methods that have been applied include 2-D wideband Multiple Signal Classification (MUSIC) algorithm, Least Square (LS), Total Least Square (TLS), Bounded Data Uncertainty (BDU) source localizers, forward-backward and dynamic programming time-delay trackers, maximum power (MP) collection beamforming, blind MP beamforming, Hidden Markov Model (H.M.M.) speech recognizer and nonlinear-dynamics signal classification methods. The applications include speaker localization and speech recognition in multimedia conference room, user localization for cellular systems, vehicle tracking and classification in the field for security surveillance, hands-free communication, hearing aids, music recording, seismic and underwater propagations, etc. The performance of the signal processing algorithms and sensor-array systems are evaluated through both computer simulations and field testing.
展开▼
