Abstract: In order to enhance the stability of the deconvolved conventional beamforming (D-CBF) algorithm, reduce background noise levels in the azimuth spectrum, and improve processing gain, a total variation constrained deconvolved CBF (TVD-CBF) spatial spectrum estimation algorithm is proposed. The approach leverages the sparse prior of the source distribution by incorporating a total variation regularization term as a nonlinear constraint within the cost function. Consequently, spatial resolution is improved while the accumulation of noise and errors is suppressed, thereby enhancing solution stability. Simulation results demonstrate that the TVD-CBF algorithm significantly outperforms the D-CBF algorithm in terms of azimuthal directivity and resolution, exhibiting surperior performance in direction of arrival estimation. The effectiveness of the TVD-CBF algorithm is further validated through experiments on sea trial data.