混合网络攻击下网络化系统的状态估计研究

Research on State Estimation of Networked Systems Under Mixed Cyber-Attack

中文关键词:  网络攻击  滤波算法  状态估计  最小方差无偏估计

英文关键词:cyber-attack  filtering algorithm  state estimation  minimum variance unbiased estimation

中文摘要:

      针对传统三步迭代滤波算法在遭受网络攻击的情况下无法对系统状态进行估计的问题，提出一种扩展三步递归滤波算法，可以对遭受混合网络攻击的网络化系统进行状态估计。分别将拒绝服务（Denial of Service,DoS)攻击建模为服从Bernoulli分布的随机变量，将虚假数据注入（False Data Injection,FDI)攻击描述为未知干扰输入项，对混合网络攻击下网络化系统的状态进行估计。通过设计新息对测量方程中的攻击信号进行估计，使用此估计值进一步修正新息，以便对系统状态进行估计。然后采用矩阵增广方式将状态方程中的攻击信号增广入系统状态值，继续设计新息对增广后的状态进行估计，此时可同时得到系统状态和2个攻击信号的估计值，且为最小方差无偏估计值。最后，使用数值仿真和无人机飞行模型仿真验证了滤波算法的有效性。

英文摘要:

      Aiming at the problem that traditional three-step iterative filtering algorithm cannot estimate the system state under cyber-attack,an extended three-step recursive filtering algorithm is proposed to estimate the state of networked systems under mixed cyber-attack.Denial of service(DoS) attacks are modeled as random variables following Bernoulli distribution and false data injection(FDI) attacks are described as unknown interference inputs,respectively,to estimate the state of networked systems under mixed cyber-attcks.By designing innovation to estimate the attack signal in the measurement equation,and using this estimated value to further modify the innovation,the system state can be estimated.Then,the attack signal in the state equation is incorporated into the system state value by using matrix augmentation method,and the innovation is further designed to estimate the expanded state.At this point,both the estimated values of the two attack signals and the system state can be obtained simultaneously,and they are the minimum variance unbiased estimates.Finally,the effectiveness of the filtering algorithm is verified through numerical simulation and unmanned aerial vehicle flight model simulation.