一种基于射频指纹的电网无线设备入侵检测算法  

An Intrusion Detection Algorithm for Power Grid Wireless Equipment Based on Radio Frequency Fingerprint

作　　者：陈子骁 周小明[2] 李钊[2] 刘晓凯 许方敏[1] CHEN Zixiao;ZHOU Xiaoming;LI Zhao;LIU Xiaokai;XU Fangmin(School of Information and Communication Engineering,Beijing University of Posts and Telecommunications,Beijing 100876,China;State Grid Liaoning Electric Power Supply Co.,Ltd.,Shenyang 110004,China;Beijing Wulian Zhitong Technology Corporation,Beijing 100000,China)

机构地区：[1]北京邮电大学信息与通信工程学院,北京100876 [2]国家电网辽宁省电力有限公司,辽宁沈阳110004 [3]北京物联智通科技有限公司,北京100000

出　　处：《无线电工程》2021年第5期352-359,共8页Radio Engineering

基　　金：复杂生物场景下的分子信号检测理论与方法研究(61971050);2020年工业互联网创新发展工程——工业企业网络安全综合防护平台。

摘　　要：随着泛在电力物联网的广泛建设,泛在电力物联网需要管理的无线终端设备日益增多,新的业务模式和复杂应用场景对电网无线设备入侵检测提出了更高的要求。射频(Radio Frequency,RF)指纹使用了终端设备射频上固有的硬件差异进行识别,且该特征无法用任何其他设备进行复制。为了解决在实际环境中遇到的设备入侵识别准确率低等问题,将射频指纹技术应用于电网无线设备入侵检测,改进了传统的基于信号双谱的特征提取方法,提出了一种提取射频信号矩形积分双谱(Square Integral Bispectra,SIB)和选择双谱的融合特征作为射频指纹特征的新方法,使用支持向量机(Support Vector Machine,SVM)作为分类器,对无线设备终端进行入侵检测。实验结果验证了将SIB和选择双谱的融合特征作为射频指纹特征来实现电网无线设备入侵检测方法的有效性和准确性。With the extensive construction of the ubiquitous power Internet of Things,the number of wireless terminal devices that need to be managed by the ubiquitous power Internet of Things is increasing.New business models and complex application scenarios have put forward higher requirements for intrusion detection of power grid wireless devices.Radio Frequency(RF)fingerprints use the inherent hardware differences in the radio frequency of the terminal equipment for recognition,and this feature cannot be copied by any other equipment.In order to solve the problem of low accuracy of equipment intrusion recognition encountered in the actual environment,the radio frequency fingerprint technology is applied to the intrusion detection of wireless equipment in the power grid,the traditional feature extraction method based on signal bispectrum is improved,and the fusion features of radio frequency signal rectangular integral bispectrum(Square Integral Bispectrum,SIB)and the selected bispectrum are extracted as the radio frequency fingerprint features,Support Vector Machine(SVM)is used to perform intrusion detection on wireless device terminals.Finally,the experimental results verify the effectiveness and accuracy of the intrusion detection method for wireless equipment in the power grid by using the fusion feature of SIB and selected bispectrum as the radio frequency fingerprint feature.

关 键 词：射频指纹 矩形积分双谱 选择双谱 支持向量机 

分 类 号：TN91[电子电信—通信与信息系统]