Abstract: Sampling matrix inversion beamforming technique(SMI) has the worse performance in the small snapshot number and high signal-noise-ratio.Diagonal loading technique can suppress noise beamforming for the small eigen value,and improve the beamforming aberration.But loading value is hard to solve.An adaptive diagonal loading beamforming algorithm is proposed in this paper.The algorithm loads adaptively diagonal matrix to enhance the algorithm robostness,according to the eigen structure of the received signal covariance matrix.Simulation results reveal that the algorithm has the better beamforming performance than conventional adaptive beamforming algorithm in the lower,middle and high signal-noise-ratio(SNR),and it has the better performance in the small snapshot number.So it has robust characteristic and the better performance.Finally the algorithm is no more complex and can be implemented easily,and it can be used widely.