Grid-free compressive beamforming using a single moving sensor of known trajectory

Abstract

Recently, the grid-free compressive sensing (GFCS) approach was proposed to perform direction of arrival (DOA) estimation of sources. With the advancement of estimation techniques using a single sensor with a known trajectory, it is proposed that a GFCS method can be extended to achieve grid- free two-dimensional localization. Through the trajectory of the sensor, the proposed approach extracts the spatial information by first reformulating the single-channel signal into multiple waveforms, where each group of consecutive waveforms satisfying the quasi-stationary condition can be constructed into a virtual array called the sub one sensor array (SOSA). The DOA of the source with respect to each SOSA is then estimated with GFCS. Accordingly, the final location of the source is computed as the point that minimizes the mean square distance to all DOA lines. Numerical and experimental results demonstrate that the proposed approach is able to perform grid-free localization of a sound source.