A Study Of Adaptive Data Beamforming And Robust Techniques

Beamforming is a key technique in array signal processing and is widely applied in radar, sonar, communication and medicine. This thesis mainly introduces the algorithms of adaptive beamforming, sidelobe cancellation and robust beamforming.Some algorithms of adaptive beamforming are given in the condition of uniform linear array, including the maximum SNR criterion, the minimum MSE criterion and the MVDR criterion, as well as LMS criterion and SMI criterion. We have analyzed the principle of each algorithm and their advantages and disadvantages, and the two main factors that affect the quality of beamforming, including array amplitude and phase error and autocorrelation matrix estimation error.The structure and algorithm of adaptive sidelobe cancellation system are introduced in this thesis. Based on the MMSE criterion and SMI, we have derived the adaptive weight vector. And another algorithm based on eigen space is given against the disadvantages of traditional sidelobe cancellation, and it can keep the sidelobe low even in the case of a small sample.We have made it clear why the waveform can’t stay robust, and introduced some kind of robust beamforming algorithms. LCMV may result in a great loss of performance, because it widens the main lobe and brings more noise. Diagonal loading is suitable for high SNR, but the amount of diagonal loading factor is not able to be defined clearly. The eigenspace based algorithm is also robust, projecting the steering vector of the desired sigal to the subspace of signal and interference. It has a perfect performance while the SNR is high enough. But when the SNR is very low this algorithm will become unsuitable, because the signal space and noise space can’t be distinguished. Worst-Case has a good performance, except that the uncertain factor of error can’t be determined. The algorithm based on Second-Order Cone Programming(SOCP) has a very good performance. It avoids equation restraints and saves System freedom, so as to enhance the system performance.In addition, we have present simulation and performance analysis for every algorithm.