Abstract: In practical application, array signals often face the interference of noise, and the performance of the beamformers decline sharply when the signal to noise ratio(SNR) is low. To solve this problem, an adaptive beamforming algorithm based on improved mutual coupling coefficient estimation is studied. According to the principle, the estimation errors of the mutual coupling coefficient under different input SNR and the estimation errors of the mutual coupling coefficient under different direction of arrival (DOA) estimation errors are studied. By analyzing the error, the original algorithm is modified to estimate the number of signals under different SNR. The proposed algorithm improves the guidance vector and the expected signal, and improves the estimation accuracy of the mutual coupling coefficient with low SNR. The comparison and analysis shows that the proposed algorithm has stronger anti-interference ability with low SNR, and has better robustness in the face of mutual coupling phenomenon.