摘要
针对信息物理系统下的虚假数据注入攻击(False Data Injection Attack, FDIA)中的随机攻击和隐蔽攻击,基于自适应卡尔曼滤波研究了攻击检测问题。常用的卡方检测可以有效检测出FDIA中的随机攻击,但是具有隐蔽性的FDIA可以绕过错误数据检测机制,使得卡方检测失败。由此在卡方检测的基础上结合相似性检测,针对系统噪声的时变特性,基于自适应卡尔曼滤波提出新的检测方法。该算法解决了实际噪声不确定性对系统的影响,且能有效检测FDIA中的随机攻击和隐蔽攻击。通过仿真验证了该方法的有效性。
Based on adaptive Kalman filter, this paper studies the problem of attack detection for random and covert attacks in false data injection attack(FDIA) in cyber-physical systems. The commonly used chi-square detection can effectively detect random attacks in FDIA, but the covert FDIA can bypass the error data detection mechanism and make chi-square detection fail. On the basis of chi-square detection, we combined similarity detection to propose a new detection method based on adaptive Kalman filter for the time-varying characteristics of system noise. The algorithm solved the influence of noise’s uncertainty on the system, and effectively detected the random and covert attacks in FDIA. The effectiveness of the proposed method was verified by simulation.
作者
王雅妮
朱翠
赵圣健
Wang Yani;Zhu Cui;Zhao Shengjian(College of Information&Communication Engineering,Beijing Information Science&Technology University,Beijing 100085,China)
出处
《计算机应用与软件》
北大核心
2022年第4期332-336,342,共6页
Computer Applications and Software
基金
国家自然科学基金项目(61603047)
北京市教委科技计划项目(KM201911232014)。
关键词
虚假数据注入攻击
自适应卡尔曼滤波
卡方检测
相似性检测
False data injection attack
Adaptive Kalman filter
Chi-square detection
Similarity detection
