基于神经网络的主瓣干扰抑制技术

A Neural Network Beamformer for Main-lobe Interference Suppression

本文讨论了基于数字波束形成的主瓣干扰的抑制问题。在已知主瓣干扰位置时,传统方法在干扰处形成零陷的同时,不能较好的实现主瓣位置的保形。本文采用神经网络算法,基于均匀线阵,通过对样本进行训练并进行处理,能够对接收信号在主瓣干扰位置形成较深的零陷,并使得零陷对主瓣的影响尽可能小,同时形成较低的旁瓣。通过仿真,当干噪比(INR)在20~30 d B时,该方法能够有效对主瓣干扰进行抑制。本文进而利用形成和差波束并进行比幅测角的方法,分析了本文所用方法对主瓣内测角S型曲线的影响,保证了对干扰进行抑制的同时,实现对目标较小误差的测角。

The suppression of main-lobe interference based on digital beamforming is discussed in this paper. The problem of forming zero-notches and keeping the shape of the main lobe cannot be achieved well at the same time by traditional methods when resented with the exact knowledge of interfering signals Direction of Arrival. In this paper,the neural network is used to form a zero notch in the interference direction of the main lobe,which based on the uniform linear array,and make the influence of the zero trap on the main lobe as small as possible,so that the influence on the S curve on the angle measurement is as small as possible. The simulation results show that the method can form a null at the interference when the interference to noise ratio( INR) is between 20 dB and 30 dB,the method can effectively suppress the main lobe interference. In this paper,the method of forming the difference beam and the angle of the amplitude measurement is used to analyze the influence of the method used in this paper on the S-curve curve of the main lobe,which ensures the suppression of the interference and achieves a small error angle measurement to the target.