| This dissertation addresses some new signal processing techniques for the impulsive noise environments. The underlying statistical models are the alpha-stable distributions. In recent years, pioneer work has been done in array processing, noise mitigation, optimal receiver design, time series modeling and many other areas using the alpha-stable theory. Impulsive processes arise in underwater, atmospheric, and other communication systems. Developing robust signal processing algorithms is of paramount importance in practice.; In the dissertation, new methods for parameter estimation, time delay and Doppler shift estimation, blind system identification and channel equalization in impulsive noise environments are studied using the alpha-stable distribution theory. First, methods for estimating the parameters (characteristic exponent and dispersion) of a symmetric alpha-stable distribution are presented. The fractional lower-order moments and their applications are introduced. These algorithms have been tested successfully using the real-world land clutter and sea clutter data. Secondly, problems of time delay estimation and joint estimation of time delay/frequency delay (Doppler shift) in impulsive noise environments are addressed. A class of new estimation algorithms are presented. These algorithms are robust in both Gaussian noise and impulsive alpha-stable noise, as demonstrated through comprehensive simulations. Thirdly, a novel approach for blind system identification based on fractional lower-order moments is proposed. The Alpha-Spectrum, a spectral representation for impulsive environments, is formulated. Conditions for blind identifiability of any FIR channel (non-minimum phase, unknown order) are established using the properties of the Alpha-Spectrum. Finally, robust technique for channel equalization is investigated. Summary and future work is discussed in conclusion. |
