基于短时傅里叶变换的无人机射频指纹分类识别  

Classification and Identification of UAV RF Fingerprints Based on Short Time Fourier Transform

作　　者：李超群 王金明 LI Chaoqun;WANG Jinming(Army Engineering University of PLA,Nanjing Jiangsu 210007,China)

机构地区：[1]中国人民解放军陆军工程大学,江苏南京210007

出　　处：《通信技术》2022年第9期1202-1207,共6页Communications Technology

摘　　要：近年来,无人机“黑飞”事件常有发生,对低空空域的安全及部分场所的安全保密带来了极大的隐患。为加强对空域无人机的识别管理,提出了一种基于短时傅里叶变换的无人机识别方法。首先对无人机遥控器射频信号进行短时傅里叶变换得到射频信号的时频谱;其次通过分析时频谱,得到能量轨迹,获取能量瞬态并提取能量瞬态特征;最后采用K-近邻算法对提取的能量瞬态特征进行分类识别,并分析比较不同分类算法的性能。分析结果表明,所提方法的识别率可达到98.19%。In recent years,UAV“black flight”events often occur,which brings great security risks to the safety of low altitude airspace and the safety and confidentiality of some places.In order to strengthen the identification management of UAV in airspace,a UAV identification method based on short-time Fourier transform is proposed.First,the time spectrum of the RF signal of the UAV remote controller is obtained through the short-time Fourier transform.Then,the energy trajectory is obtained through time spectrum analysis.The energy transient is obtained,and the energy transient features are extracted.Finally,the extracted energy transient features are classified and identified by the K-nearest neighbor algorithm,and the performance of different classification algorithms is analyzed and compared.The analysis results indicate that the classification accuracy of the proposed method is 98.19%.

关 键 词：短时傅里叶变换 能量瞬态 特征提取 射频指纹识别 

分 类 号：TN911.7[电子电信—通信与信息系统]