Reserach On Robust Adaptive Beamforming Algrithms

Adaptive beamforming is one of the main research directions in array signal processing,which utilizes a plurality of sensor arrays of different positions in the space for signal acquisition.The signal is optimally received by adjusting the weighting coefficients of the elements in the array,and adaptive beamforming is widely used in radar antennas,navigation systems,and smart home appliances.However,there are various errors in the actual engineering application.the existence of these errors seriously affect the performance of the adaptive beamforming algorithm.Therefore,it is necessary to study the robustness of the adaptive beamforming algorithm in the presence of array errors.In this paper,we use the NAMI(Noise-Alone Matrix Inverse)method and the SPNMI(Signal-Plus-Noise Matrix Inverse)method to analyze the array error under the minimum variance distortionless response.The results show that the sampled data contains the expected signal is more sensitive to array errors compared with the case without the expected signal;The performance of the traditional robust beamforming algorithm is analyzed in the presence of array errors.It is concluded that the diagonal loading algorithm is too dependent on the loading factor and there is no definite method to determine the loading factor,the feature subspace algorithm requires the number of known signal sources and interference sources and is only suitable for the case of strong power expectation signals and the robust Capon beamforming is too complicated.At the same time,the robustness of the traditional algorithm is difficult to maintain under the condition of low input signal-to-noise ratio and small number of sampling snapshots;Aiming at the existing problems of traditional algorithms,an improved algorithm is proposed based on the characteristic subspace principle.The improved algorithm can reduce the influence of noise by reconstructing the covariance matrix to make the signal subspace more accurately.At the same time,unitary transformation is introduced to convert matrix operations to real numbers,which reduces the difficulty of matrices in the process of inverse and eigenvalue decomposition,and reduces the computational complexity of the algorithm to a certain degree.The correctness and robustness of the improved algorithm are verified by simulation experiments under low SNR and low sampling snapshots.