A Bayesian Compressive Sensing strategy for direction-of-arrival estimation
Matteo Carlin, Paolo Rocca · 2012
An innovative approach for the real-time direction-of-arrival (DoA) estimation of multiple signals impinging on a linear array is presented. Starting from a Bayesian Compressive Sensing formulation of the DoA detection problem, the proposed methodology searches for the most likely directions for the impinging signals and provides a “confidence level” for the obtained solution. Towards this end, the data acquired from the array sensors are processed through a numerically-efficient Relevance Vector Machine. A set of representative numerical results, concerned with both single and multiple signals, is provided to preliminarily assess the features and advantages of the proposed technique.