Beamspace Eigenstructure-based Algorithm And Low Sidelobe Implementation For Adaptive Beamforming

Beam-space processing method inherits and learns from some of the array-space methods, which is a development and innovation of array-space processing.Currently, with continuous in-depth study, more and more scholars are increasingly interested in beam-space adaptive beamforming. Many algorithms have been applied to beam-space, and their performance in the beam-space is also studied; Eigen-space method is one of the methods commonly used in array-space, the beamformer based on these methods has a good convergence speed and robustness, by combining the beam-space and eigen-space processing methods to obtain a beamformer with both beam-space and eigen-space properties is a research subject worthy of study.This paper firstly introduces the beam-space and eigen-space processing methods researth progress at home and abroad, there are many array-space methods having been successfully applied in beam-space, which has greatly improved the performance of the beamformer; The eigen-space method has an excellent performance in array-space, algorithms based on eigen-space are also countless.Secondly, basic knowledge related to array signal processing is introduced, including the basic principles of the array and the array beamforming. Spatial signal and its modulation-demodulation are introduced in the part of array basic principles; the output SNR from beamformer is analyzed in the part of array beamforming.Then, a summary of algorithms for beamforming is maked, containing two major parts:the first part is the general principles of beamforming, which includes the conventional beamforming and optimum beamforming; the second part is the general implementation methods of adaptive beamforming, which includes adaptive weight and adaptive sampling matrix.Finally, a new beamspace beamforming algorithm is proposed, the research work includes two major parts:Application of the eigen-space method in the beamspace is proposed in the first part, including the conversion of array signals, the beamspace processing under the criterion of linear constrained minimum variance, the transform matrix and so on. By comparing the beam patterns generated by beamspace eigen-based method and the conventional method, analysis about the new algorithm is proposed in terms of the robustness and convergence performance, the ratio of main and side lobe level, the gain of the array output signal to interference plus noise ratio, the required number of array samples and so on.Three methods of getting beam-space patterns are proposed for low-sidelobe in the second part. The simulation results of the first method show that the directly Window taper method on array weight vector is no longer suitable for beam-space processing; the simulation results of the second method show that Window taper method on beamspace transform matrix T is applicable to the case that the input signal SNR is relatively greater than input INR, when the input signal SNR is relatively smaller than the input INR, the sidelobe level obtained is not ideal; the simulation results of the third method show that beam pattern sidelobe obtained from the virtual-interference-method changes with the changes of virtual-interference INR envelope, a ideal level of side lobe can be obtained by changing virtual-interference INR envelope through programming, which is an ideal way to reduce the sidelobe.