Enhancement of two-dimensional acoustic source identification with Fourier-based deconvolution beamforming

Abstract

When utilizing conventional regular focus point distributions to define a relatively large source region, Fourier-based deconvolution beamforming, an attractive acoustic source identification technique, would suffer from some limitations: 1) significantly deteriorative location and quantification accuracy for sources away from the center of the focus region; 2) pronounced sidelobe contaminations. The arch-criminal is the assumption that the point spread function (PSF) is definitely shift-invariant over the entire focus region fails to be satisfied well. This paper focuses on remedying these limitations for two-dimensional (2D) acoustic source identification. First and foremost, a novel focus point generation approach is introduced, which can generate unconventional irregular 2D focus point distributions tending to make PSF more shift-invariant. Additionally, a sidelobe suppression approach is suggested. Effects of these approaches are examined both with computer simulations and experimentally. This study provides the feasibility of using Fourier-based deconvolution beamforming to accurately and efficiently identify acoustic sources in a relatively large region.