海军航空大学学报2025，Vol.40Issue(4)：547-556,10.DOI:10.7682/j.issn.2097-1427.2025.04.005
基于幅度概率分布特征的无人机检测识别方法
UAV Detection and Recognition Method Based on RF Signal Amplitude Distribution Characteristics
摘要
Abstract
Unmanned aerial vehicle(UAV)technology is booming,which is widely used in many fields such as remote sensing and mapping,meteorological monitoring,agriculture and forestry plant protection,industrial safety,geological ex-ploration,etc.However,the widespread application of UAV has also brought new challenges to safety management and monitoring.Therefore,the new detection and identification technology for UAV has gradually become a research hotspot.Fortunately,the UAV detection based on radio frequency(RF)signal is an effective means to achieve the UAV supervi-sion.In order to improve the detection and recognition probability in low signal-to-noise ratio(SNR),a UAV detection and recognition method based on the amplitude distribution characteristics of RF signal is proposed.Firstly,the original RF signal is preprocessed by the adaptive triangular threshold method to re-move the invalid signal region in the sampled signal and reduce the processing complexity.Then,the amplitude interval probability distribution and constellation circle interval probability distribution of the preprocessed signals are extracted to train the classification network.Finally,a sup-port vector machine(SVM)network is designed to detect the existence of the UAV in real time.The algorithm is verified on the open data set,and the experimental results show that the detection rate is above 98%when the SNR is larger than-5 dB.The proposed method enhances the detection rate by 20.71%at a low SNR of-30 dB to-5 dB,surpassing the per-formance of the detection methods for time-domain and frequency-domain features.关键词
无人机/检测识别方法/射频指纹/信号预处理/幅度概率分布/星座圆间隔概率分布Key words
UAV/detection and recognition method/RF fingerprint/signal preprocessing/amplitude interval probability distribution/constellation circle interval probability distribution分类
航空航天引用本文复制引用
陈淑婷,刘子悦,付元华,杜鹏飞,杨伟..基于幅度概率分布特征的无人机检测识别方法[J].海军航空大学学报,2025,40(4):547-556,10.基金项目
国家自然科学基金(62301447) （62301447）
四川省自然科学基金(2023NSFSC1377) （2023NSFSC1377）
