Improvement of Fourier-based fast iterative shrinkage-thresholding deconvolution algorithm for acoustic source identification

Abstract

Fourier-based fast iterative shrinkage-thresholding algorithm (FFT-FISTA) is a novel beamforming deconvolution method for acoustic source identification. Regrettably, there is a deficiency of failing to identify the acoustic source far from the center of calculation plane, when utilizing Fourier-based beamforming deconvolution method with the traditional regular focus point grid. This paper devotes to overcoming the deficiency by a new irregular focus point grid. Simulations and validation experiment both show that FFT-FISTA with an irregular focus point grid can overcome the forementioned deficiency. Besides, it is superior to the FFT-FISTA with a regular focus point grid in terms of spatial resolution improvement, sidelobe removement and source quantification, when the acoustic source is close to the center of calculation plane. This work contributes to enlarge the acoustic source identification area of FFT-FISTA and improve its performance.