| Many theories of mathematical statistics and time series can find wide applications in signal processing and communications. This thesis focuses on an important area of signal processing, namely, statistical signal processing in the impulsive noise environments. Such environments arise in underwater, atmospheric, and other mobile communications. Developing robust algorithms for such scenarios is of paramount importance in many engineering fields.; In this thesis new methods for parameter estimation, blind system identification and adaptive equalization for impulsive noise environments are presented. An important application in radio network using the stable distribution theory is also addressed. The underlying statistical model is the symmetric alpha-stable distribution. First, methods for estimating the parameters (characteristic exponent and dispersion) of a symmetric alpha-stable distribution are presented and applied to real-world sonar clutter data modeling. Secondly, a new algorithm for blind channel 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 channels (non-minimum phase, unknown order) are established using the properties of the Alpha-Spectrum. Thirdly, the robustness issue of adaptive equalization technique is investigated. Finally, the modeling of interference in a wireless communication network is studied and shown to have a stable distribution. Summary and future work are included in the conclusion. |
