Source localization using acoustic vector sensor

Direction-of-arrival (DOA) estimation has extensive applications in the field of signal pro-cessing. Most of the work on DOA estimation is done using an array of sensors. But this arrangement is computationally expensive and space consuming. These problems are overcome by using acoustic vector sensors (AVS). In a real world scenario, the speech signal (where direction has to be estimated) is corrupted by noise, signal reflection and re-verberation. Hence, an estimate of the direction is less accurate. In this dissertation work, existing algorithms for DOA estimation using AVS are discussed. The limitations of these algorithms are analysed and better and more general methods are introduced for DOA es-timation. The work also includes algorithms to denoise the noisy speech signal so that the estimation can be made more accurate. Various algorithms are discussed, compared and analysed using various quantifying parameters for both noisy and denoised speech using MATLAB as simulation tool.Master of Science (Signal Processing