Low-Complexity Variable Loading For Robust Adaptive Beamforming

Adaptive beamforming has an important place in array signal processing. It is widely applied in wireless communication, radar, sonar, microphone array speech processing, medical imaging, radio astronomy and other fields. The adaptive beamforming is designed to enhance the desired signal while suppress interferences and noise. In practical applications, due to the errors in steering vector and covariance matrix, the improper modeling or variations in communication environment, the adaptive beamformer suffers significant performance degradation. Therefore, it is necessary to do some researches on the robust beamforming techniques.In recent decades, researchers have proposed many robust adaptive beamforming algorithm, and the diagonal loading is one of the important algorithms. In this thesis, we propose a new low complexity variable loading algorithm, which is based on the traditional variable loading algorithm and the ad hoc diagonal loading algorithm.There are two innovations in the proposed algorithm:1) we use the maximum likelihood method to modify the eigenvalues of sample covariance matrix. When the eigenvalue is smaller than noise power σn2, we replace the eigenvalue with σn2, so that we can reduce the noise disturbance of small eigenvalues when the snapshot is small.2) we apply the loading factor of the ad hoc diagonal loading algorithm to the traditional variable loading algorithm. The loading factors of small eigenvalues will be 10σn2, which can enhance the robustness of the beamformer. While for the bigger eigenvalues which correspond to the signal subspace, the loading factors will be much smaller than 10σn2,which will enhance the ability to suppress interferences. And the primary complexity in our algorithm is the eigenvalue decomposition of matrix. The main content of this thesis is summarized as follows:(1) Introduce the research background of beamforming techniques together with the basic knowledges, common optimization methods, and measure criterions of adaptive beamforming techniques, and analyze the performance of Capon beamformer.(2) Introduce several classic robust beamforming techniques, summarize their research background, and then derive their computing process in detail, and analyze their performances with simulation experiments.(3) Analyze the advantages and disadvantages of the traditional variable loading algorithm and the ad hoc diagonal loading algorithm, then creatively combine the advantages of two algorithms and propose a new low complexity variable loading algorithm. Verify the rationality of the maximum likelihood estimation algorithm. Compare the proposed algorithm with the robust algorithms which are above mentioned, and verify its performance by several simulation experiments. The simulation results show that the performances of the proposed algorithm are better than those of other robust algorithms. In order to reflect the influence of eigenvalues modifications on the algorithm of the thesis, we compare the performance of the proposed algorithm and the algorithm without eigenvalues modifications. At last, according to the simulation results, we look forward the future research works.