一种基于神经网络的自适应波束形成方法  

An Optimum Beamforming Algorithm Based on Neural Network

作　　者：吴甜[1] 崔宏宇 WU Tian;CUI Hongyu(Nanjing Research Institute of Electronics Tech0nology,Nanjing 210039,China)

机构地区：[1]南京电子技术研究所,南京210039

出　　处：《移动信息》2024年第7期257-259,共3页MOBILE INFORMATION

摘　　要：自适应波束形成是数字智能天线中的关键技术,具有主瓣灵活、超低副瓣、干扰零陷等优点。文中提出了一种基于神经网络的智能波束形成方法,该方法以相对于期望输出信号的线性约束最小方差准则(LCMV)为收敛条件,能根据外界环境自适应调整天线权值,利用神经网络非线性和并行计算的特点,克服传统自适应波束形成方法计算复杂和收敛速度慢的问题。仿真实验验证了该方法的有效性以及其在不同外界环境下的适应能力。Adaptive beamforming is a key technology in digital smart antennas,which has the advantages of flexible main lobe,ultra-low side lobe,and zero interference notch.In this paper,an intelligent beamforming method based on neural networks is proposed.The method takes the linear constraint minimum variance criterion(LCMV)relative to the expected output signal as the convergence condition,and can adaptively adjust the antenna weights according to the external environment.The characteristics of neural networks nonlinearity and parallel computing are used to overcome the problems of complicated calculation and slow convergence of traditional adaptive beamforming methods.Simulation experiments verify the effectiveness of the method and its adaptability to different external environments.

关 键 词：波束形成 智能天线 神经网络 

分 类 号：TN82[电子电信—信息与通信工程]