基于遥控信号频谱特征的无人机识别算法

UAV Recognition Algorithm Based on Spectrum Characteristics of Remote Control Signal

近年来,无人机"黑飞"事件与非法入侵的案例开始激增,现有的识别技术往往难以实时、准确地检测到非法无人机,为此提出了一种基于遥控信号频谱特征的无人机识别算法。该算法通过背景频谱模板学习(利用二次平均的方法求信号检测阈值,并屏蔽宽带信号对无人机遥控信号检测的干扰)、无人机跳频与定频遥控信号检测、机型判断及其特征参数学习三大模块来识别非法入侵的无人机。实验结果表明,该算法可以实现对常见无人机实时准确地识别,且鲁棒性较强。

In recent years,the cases of“black flying”and illegal invasion of UAVs have been increasing rapidly.It is often difficult for existing identification techniques to detect UAVs accurately in real time.A UAV recognition algorithm based on the spectrum characteristics of remote control signals is proposed.The algorithm uses three modules to identify the illegal invading UAVs,which include background spectrum template learning(using the double-averaging method to obtain signal detection threshold,and shielding the interference of wideband signal to UAV remote control signal detection),UAV frequency hopping and fixed frequency remote control signal detection,and model judgment and its characteristic parameters learning.Many experiments show that the algorithm can realize real-time and accurate recognition of common UAVs and has strong robustness.