Abstract: Due to excessive computation of conventional deconvolution beamforming algorithm in processing wideband random signals, a fast deconvolution beamforming method conductive to engineering applications is proposed. According to the similarity of the beam patterns at different frequencies, the wideband random signal is divided into multiple narrow bands with non-equal spacing, and the approximate processing that takes the point spread function (PSF) at the central frequency point in each narrow band for deconvolution is performed, so that the calculation speed of deconvolution beamforming can be greatly improved. By expanding the boundary on the beam power spectrum, the boundary blur problem caused by the Richardson-Lucy (R-L) iterative algorithm is solved, and the calculation speed is further improved. Simulation and sea trail results confirm that this method has higher resolution, higher processing gain and better sidelobe suppression compared to conventional beamforming; and the operational speed is increased by more than 50% compared with conventional deconvolution beamforming method.