Research On Robust Anti-interference And Beamforming Algorithm

Because of the broad application of adaptive beamforming technology in various fields, the robustness of adaptive beamforming has been one of the research focuses of the array signal processing. However, when there is the desired signal in the training data, most of classical adaptive beamforming algorithms are adversely affected by the power and the steering vector mismatch of desired signal. Especially for the case of the strong desired signal, the robustness of the algorithm needs to be improved. In recent decades, domestic and international researchers proposed many methods to improve the robustness of adaptive beamforming. Based on the presence of two kind of the steering vector mismatches, cause by only the DOA mismatch of the desired signal and the amplitude and phase errors of channels, respectively, the performance of classical and improved adaptive beamforming algorithms are deeply studied in this dissertation.Some basic knowledge of the array signal processing and several classical adaptive beamforming algorithms are introduced in this dissertation at first. Via simulation experiment, their performance, affected by desired signal power and the snapshot, are analyzed based on the above two kind of steering vector mismatches, respectively.Simulation results demonstrate that the Capon beamformer will have a serious performance decline under the non-ideality factor mentioned above. Even though diagonal loading beamformer and eigenspace-based beamformer can improve the performance in a certain extent, they have poor performance in the presence of the steering vector mismatch when the desired signal is strong.To improve the poor performance of classical adaptive beamforming algorithms in the presence of the strong desired signal, some novel robust adaptive beamforming algorithms based on covariance matrix reconstruction are researched. By trying to remove the desired signal component from sample covariance matrix and estimating a new steering vector of desired signal, these novel algorithms can improve the robustness of these algorithms in the presence of the steering vector mismatch with strong desired signal.There are several algorithms which can improve the ability to resist DOA mismatch by correcting steering vector are researched in this dissertation. Based on the knowledge of a space range that the DOA of the desired signal located, the actualsteering vector of desired signal can be estimated by using the quadratic programming and the space projection method. Simulation results show that these algorithm can perform very well on steering vector mismatch.A new beamforming algorithm based on the structure of side-lobe cancellation and space projection method is proposed in this dissertation at last. By trying to maximize the desired signal power, a corrected steering vector can be obtained, which can improve the robustness of beamforming. Compared with the side-lobe cancellation algorithm, it improves the robustness in the presence of the steering vector mismatch.And compared with the space projection method, the new algorithms can reduce the complexity of the algorithm without any loss of performance.