| Array signal processing has been proven to be a successful technology in many applications. In recent years, there has been an increasing interest in applying such discipline in the sensor network research. In this dissertation, we focus on the theory and techniques of wideband source localization using arrays of passive sensors (e.g., acoustic and seismic). We propose a maximum-likelihood estimator, which is optimized in a single step as opposed to other estimators that are optimized separately in relative time-delay and source location estimations. For the multi-source case, we propose an efficient alternating projection procedure based on sequential iterative search on single source parameters. The proposed algorithm is shown to yield superior performance over other suboptimal techniques and is efficient with respect to the derived Cramér-Rao bound.; The theoretical Cramér-Rao bounds are derived and analyzed for wideband source localization and DOA estimation. The Cramér-Rao bound analysis shows that better estimates can be obtained for high frequency signals than low frequency signals, and when the source signal is unknown, large range estimation results but the angle estimation is unaffected. The Cramér-Rao bound also suggests the uniformly spaced circular array as a favorable geometry for most cases.; This dissertation also addresses other practical issues that are crucial for source localization, such as determining the number of sources, estimating unknown sensor locations, and estimating the reverberated channel response. To determine the number of sources in the data, we propose a sequential detection scheme based on a simple statistical test along with the ML parameter estimation in each step. Then, we extend the ML estimator to estimate the range from a source to an unknown sensor location and formulate a least squares estimator for the uncalibrated sensor location. At last, we also consider a training based acoustic channel estimation technique with interference cancellation. For the proposed algorithms in this dissertation, extensive computer simulations and experimental results are presented to demonstrate the usefulness of these algorithms. |
