Compression computational grid based on functional beamforming for acoustic source localization

Phased microphone arrays have become a standard technique for acoustic source localization. Compared with beamforming algorithms such as the conventional beamforming, deconvolution approaches such as DAMAS successfully improve the spatial resolution. However deconvolution approaches usually require high computational effort compared to beamforming algorithms. Without optimizing deconvolution algorithm, recently DAMAS with compression computational grid based on the conventional beamforming (denoted by DAMAS-CG2) has reduced computational run time of DAMAS in applications (Ma and Liu, 2017). This paper proposes a novel algorithm that DAMAS with a novel compression computational grid based on an advance beamforming algorithm functional beamforming (denoted by DAMAS-CG3). This new algorithm takes advantages of functional beamforming to obtain large compression ratio. Simulated applications and experimental applications of benchmark test DLR1 show that DAMAS-CG3 is one order of magnitude faster than DAMAS-CG2 in most cases. In addition, the advantage of DAMAS-CG3 compared to DAMAS-CG2 is particularly more obvious with the threshold decreasing. However for some extreme situations that very complicated sources distribute to a larger extent relative to the scanning plane, the advantage of DAMAS-CG3 compared to DAMAS-CG2 may disappear. In order to get a large compression ratio in any application, the authors highly recommend compressing computational grid based on not only conventional beamforming but also functional beamforming, and then choosing the compression grid with larger compression ratio.