Periodic boundary based FFT-FISTA for sound source identification

Abstract

Compared with the conventional beamforming, the Fourier-based fast iterative shrinkage thresholding algorithm (FFT-FISTA) can effectively improve the spatial resolution and suppress the sidelobe. To furtherly achieve higher computational efficiency and better sound source identification performance, an alternative periodic boundary is utilized to replace the zero boundary of Fourier transform, a periodic boundary based FFT-FISTA is proposed in this paper. And its superiority is demonstrated by the simulation and validation experiment of equal and unequal intensity sources.