Research On Adaptive Beamforming Based On Direction Of Arrival Estimation

This dissertation is focusing on array signal processing, laying emphasis on adaptive beamforming based on direction-of-arrival estimation, especially on diagonal loading and linear constraining techniques for performance improvement and robustness enhancement of adaptive array beamformers.First, the principle of classical MUSIC algorithm of spatial spectrum estimation is discussed with performance comparison by simulation. After that, the adaptive beamforming algorithms based on diagonal loading are discussed in detail, and an improved adaptive beamforming algorithm suitable for the scenario of limited snapshot number is proposed, which could dynamically adjust the diagonal loading values according to the array data by the use of the relationship between the diagonal loading values and the estimation error of the covariance matrix, and obtained better performance than the conventional algorithm based on fixed diagonal loading value. Normally, the constant modulus algorithm converged to the signal with greatest power among the cochannel signals received by the array. Based on a study on CM algorithm a LCMV guided constant modulus algorithm is proposed and discussed in detail in this paper, which has output signal with steady SINR and converges fast even when the interfering signal is stronger than the desired signal. Finally, an algorithm for the blind separation and recovery of correlated cochannel sources based on the constant modulus array is provided, which performs well even for highly correlated sources and without performance loss experienced by the cascade system.Theoretical analysis and mathematical derivations for different algorithms are provided along with explaining of corresponding application prerequisites. Numerous simulation results have proved the feasibility and effectiveness of all the proposed algorithms in this paper.