Norm-constraining beamforming amid under-ice non-Gaussian noise

A norm-constraining beamforming method is proposed to solve the problem of large errors in bearing estimation due to noise model mismatch in the under-ice environment. The norm is applied on the received signal to optimize the data covariance matrix, overcome the model mismatch caused by non-Gaussian noise and improve the accuracy of bearing estimation. The analysis of the norm constraining performance shows that the constraining processing can efficiently suppress the non-Gaussian noise. The results of simulation and experiment show that, applying beamforming after norm constraint of the received signal in the under-ice environment can effectively improve the accuracy and stability of the bearing estimation.