基于车辆环境态势感知的毫米波波束跟踪  

Millimeter Wave Beam Tracking Based on Vehicle Environmental Situational Awareness

作　　者：张璐璐 仲伟志[1] 张俊杰 朱秋明[2] 陈小敏[2] ZHANG Lulu;ZHONG Weizhi;ZHANG Junjie;ZHU Qiuming;CHEN Xiaomin(College of Astronautics,Nanjing University of Aeronautics and Astronautics,Nanjing,Jiangsu 210016,China;College of Electronic and Information Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing,Jiangsu 211106,China)

机构地区：[1]南京航空航天大学航天学院,江苏南京210016 [2]南京航空航天大学电子信息工程学院,江苏南京211106

出　　处：《信号处理》2022年第3期457-465,共9页Journal of Signal Processing

基　　金：国家自然科学基金重大仪器研制项目,无人机频谱认知仪(61827801)。

摘　　要：针对以往车联网动态环境下采用的波束搜索方法无法解决毫米波窄波束实时匹配这一问题,本文结合射线追踪法,提出采用一种基于车辆环境态势感知的波束搜索方法,以提高波束搜索效率。该方法首先通过射线追踪法计算得到最优波束对指数,并对当前场景进行编码,形成特征向量,建立数据库;而后采用机器学习对测试车辆场景进行训练,产生适用于实际情况下车辆网波束匹配的机器学习模型;最后在该模型下进行实际目标车辆的态势搜索,仿真结果表明,该方法能够有效降低波束搜索开销,在保证波束搜索精度的同时提高波束搜索效率。To address the problem that the previous beam search methods used in the dynamic environment of Telematics cannot solve the problem of real-time matching of millimeter wave narrow beams,this paper proposes a beam search method based on the situational awareness of the vehicle environment in combination with ray tracing method to improve the efficiency of beam search. The method first calculates the optimal beam pair index by ray tracing and encodes the current scene to form a feature vector and build a database;then machine learning is used to train the test vehicle scene to generate a machine learning model for matching the beam of the vehicle network in real situations;Finally a situational search for the actual target vehicle is carried out under the model. The simulation results show that the method can effectively reduce the beam search overhead,improve beam search efficiency while ensuring beam search accuracy.

关 键 词：车联网 毫米波 机器学习 态势感知 

分 类 号：TN928[电子电信—通信与信息系统]