Blind Beamforming And Doa Estimation

Blind beamforming(BBF) is concerned with the reconstruction of source signals from the outputs of a sensor array without a priori information of the direction of the desired source, relying instead on signal characteristics and array structures. In order to be applicable to practical underwater systems, several kinds of the blind beamforming algorithms were discussed in this thesis, and direction-of-arrival (DOA) estimation based on BBF is also discussed.In the thesis, three kinds of blind beamforming algorithms were introduced, which are the cumulant-bascd blind beamforming, the array structure-based blind beamforming, and the constant modulus algorithm. In the constant modulus algorithm, the multistage CM (constant modulus) array and the analytical constant modulus algorithm were analyzed. Combining the blind beamforming with the estimation of DOA, the DOA can be estimated at the same time when the source signals are recovered.Based on theoretical studies on these algorithms, by means of computer simulation and the water tank experiment, we analyzed the performance of these algorithms in uniform linear array, which included the separation and reconstruction of source signals and DOA estimation performance. The results showed that the blind beamforming algorithms could recover the desired source and null out the interfering sources. When the input SNR was higher than the SNR threshold, these algorithms could correctly estimate the DOA. And the algorithms in thesis were robust to the model error of array geometry. With higher input SNR, the angle separation ability of these algorithms were beyond the Rayleigh limitation. However, due to the difference in information utilized in various algorithms, they are suitable to different situations and have different abilities in rejecting interference and computational burdens.