复杂环境下高扩展性无人机信号检测识别方法  

UAV Signal Detection and Identification Method with High-Expansibility in Complex Environment

作　　者：葛嘉鑫 李晋徽 晋晓曦 张涵硕 温志津 GE Jiaxin;LI Jinhui;JIN Xiaoxi;ZHANG Hanshuo;WEN Zhijin(Academy of Military Sciences,Beijing 100191,China;Xidian University,Xi’an Shaanxi 710126,China)

机构地区：[1]军事科学院,北京100191 [2]西安电子科技大学,陕西西安710126

出　　处：《通信技术》2023年第8期929-936,共8页Communications Technology

基　　金：国防科技和军事理论卓越青年科学基金(2022-JCJQ-ZQ-024)。

摘　　要：利用无人机与遥控器间的无线电信号进行无人机检测和识别是当前的研究热点。但在该领域,仍存在两个关键问题有待解决:一是如何在存在众多第三方信号的复杂电磁环境下有效检测识别无人机信号,二是如何保证检测识别系统针对新型无人机的快速扩展能力。针对这两个问题,提出了一种高扩展性的无人机信号检测识别架构。首先,利用YOLO神经网络模型应对复杂电磁环境下信号检测难题。该模型设计和训练面向通用电磁信号检测。完成信号检测后,利用“信号特征提取+支持向量机”结构设计无人机信号识别算法。该步骤计算复杂度低,模型参数少,因此对新型无人机具备良好的可扩展性。The use of radio frequency signals between UAVs and remote controllers for UAV detection and identification is a hot research topic at present.However,there are still two key problems to be addressed in this field:One is how to effectively detect and identify UAV signals in a complex electromagnetic environment with many third-party signals,and the other is how to ensure the rapid expansion of the detection and identification system for new UAVs.Aiming at these two problems,this paper proposes a highly expanded UAV signal detection and identification architecture.First,the YOLO neural network model is used to solve the problem of signal detection in complex electromagnetic environments.The model is designed and trained for general electromagnetic signal detection.Then,after completing the signal detection,the UAV signal identification algorithm is designed by using the structure of“signal feature extraction+support vector machine”.This step has low computational complexity and few model parameters,so it has good expansibility for new UAVs.

关 键 词：无人机信号检测识别 复杂环境 高扩展性 YOLO 支持向量机 

分 类 号：TN971[电子电信—信号与信息处理]