Parameter estimation for spatially-spread sources

This thesis attempts to discuss the problem of parameter estimation of spatially-spread sources. Traditional estimation techniques for those parameters have been designed only for point source which travels along a single path to an antenna array. However, in practical applications such as wireless communications, multipath propagation due to local scattering of the sources can cause spreading of the signal energy around nominal direction-of-arrival (DOA). Therefore, the traditional estimation techniques which are based on point source assumption may result in serious performance degradation.; There exist certain algorithms for spatially-spread source parameter estimation. In the process of researching and writing this thesis, the author has conducted literature reviews and has proposed a noval algorithm called Spread Minimum Variance Distortionless Response (SMVDR) algorithm under a parametric distributional model. There are two parameters under the model: nominal direction-of-arrival and angular spread. The algorithm is more computationally efficient. Empirical results have demonstrated an improved performance of the new technique. Theoretical analysis of SMVDR performance is also given in this thesis.