| 摘要: |
| 随着民用无人机的普及，无人机“黑飞”事件频频发生，给公共安全带来极大隐患。为了实现对“黑飞”无人机的有效监管，通过提取遥控信号指纹特征对无人机识别是一种有效的方法。基于民用小型无人机遥控信号通常采用跳频通信这一特性，通过分形贝叶斯变点检测算法对实测无人机遥控信号的瞬态起始点进行检测，并提取信号瞬态部分所含有的指纹特征，由主成分分析法进行特征降维，最后采用多分类支持向量算法对该信号进行分类及识别。实验结果表明，采用射频指纹法能够完成无人机型号的区分以及同一型号无人机的区分。 |
| 关键词: 民用小型无人机 射频指纹 遥控信号；分类识别 分形贝叶斯变点检测 |
| DOI： |
|
| 基金项目:广西科技重大专项(桂科AA17202022)；认知无线电与信息处理教育部重点实验室主任基金项目（CRKL180105）;广西研究生教育创新计划项目（2020YCXS021）；桂林电子科技大学研究生优秀学位论文培育项目资助（18YJPYSS07） |
|
| A radio frequency fingerprint identification method for civil small UAVs |
| JIANG Ping,XIE Yuelei |
| (Research Center for Wideband and Intelligence Information Technology,Guilin University of Electronic Technology,Guilin 541004,China) |
| Abstract: |
| With the popularization of unmanned aerial vehicles(UAVs),the illegal incident of UAV happens frequently,which brings a significant threat to public security.In order to achieve effective supervision for illegal UAV,it is an effective method to extract fingerprint features of remote control signals for UAV identification.Based on the characteristic that frequency hopping communication is usually used in the remote control signal of civil small UAV,this paper uses fractal Bayesian change point detection algorithm to detect the transient starting point of the measured UAV remote control signal,and extracts the fingerprint features contained in the transient part of the signal.The feature dimension is reduced by principal component analysis（PCA）.Finally,the multi-classification support vector machine（SVM） algorithm is used to classify the signal.The experimental results show that the radio frequency distinct native attribute(RF-DNA) method can be used to distinguish the UAV model and even the same type UAV. |
| Key words: civil small UAV RF-DNA remote control signal classification and recognition fractal Bayesian change point detection |
