Investigation of Sparse Bayesian Learning based Direction of Arrival Estimation Algorithms for a Linear Array

Abstract

In the field of statistical signal processing, Direction of Arrival (DOA) Estimation is newlineone of the major problems of interest which has become a hot topic in the recent years newlineof signal processing research. DOA estimation is a process of retrieving the newlinedirection/angle information of one or more incoming electromagnetic signals at the newlinereceiver end, which comprises of a Uniform Linear Array (ULA). The signal s newlinedirection/angle is estimated by processing the signals which are received by this newlineUniform Linear Array (ULA). The problem of DOA estimation can be reconsidered newlineas the received signal s angle estimation from the help of received signal itself. The newlinedirection/angle from which the signals are received at the receiver ULA is hidden newlineinside the ULA received signal amplitudes, which can be processed to extract the newlinedirection information using various statistical and sparse based estimation algorithms. newlineDOA estimation finds its importance in many application areas such as RADAR, newlineSONAR, wireless communication, seismology, exploration, astronomy and newlinebiomedicine as in these applications the direction of target, seismic source and mobile newlineusers are essential. The research work presented in this thesis mainly focuses on DOA newlineestimation problem using several sparse based signal processing algorithms. In newlineparticularly, Sparse Bayesian Learning (SBL) technique along with the Expectation newlineMaximization (EM) framework is developed, modified, applied and analysed for newlinedifferent performance evaluation criteria newline