Robust Acoustic Source Localization Method Based on Compressed Sensing

This paper applies the theory of Compressed Sensing to the acoustic source localization with the microphone arrays and develops a new high precision and robust acoustic source localization method. The proposed method firstly compresses the signal with a random compressed matrix and then estimates the directions of arrival by OMP algorithm. The simulation results show that, compared with the traditional MUSIC algorithm, our algorithm only needs far fewer snapshots for high-precision DOA estimation under the premise of an unknown number of acoustic target. Our algorithm has significant advantages such as super-resolution, high-accuracy, low-latency, robustness to noise and limited data, robustness to correlation of the acoustic sources etc.