Abstract:In cyber–physical systems (CPSs), secure estimation observes the states of a physical plant remotely with potentially attacked data. Most attacks in CPSs are carefully designed such that they can not only ruin the estimation but also bypass the detector. These attacked estimations would corrupt decision-making and control, leading to decreased production efficiency and even equipment damage. Therefore, detecting these attacks is of great importance. It is noted that one reason these attacks can be successfully launched is that attackers can inject arbitrary false data into the communication channel at any time step, which makes it possible that the transmission data is tampered with by any designed attack sequences. In this paper, a modified data transmission mechanism is deployed to limit the attacker's ability, and attack detectors are designed for detecting tampering attacks. Specifically, first of all, a one-bit event-triggered mechanism is deployed to binarize the transmitted data. Secondly, a primary detector is deployed to detect non-binary data received by the observer, and hence the attacker's ability is greatly reduced. Thirdly, a main detector is designed, which definitely detects tampering attacks that can bypass the primary detector.The necessary and sufficient conditions for definite attack detection are given, with which the gain of interval observable and the threshold of the event-trigger mechanism are designed. Finally, simulation results are presented to verify the analysis.

Optimal strategy of data tampering attacks for FIR system identification with average entropy and binary‐valued observations

Optimal attack strategy in system identification with binary‐valued observations: An information entropy‐based approach

State estimation for linear discrete-time systems with binary-valued quantized innovations against data tampering attacks

Binary Observation‐Based FIR System Identification Against Replay Attacks

Stealthy Actuator Signal Attacks in Stochastic Control Systems: Performance and Limitations.

Binary observation‐based identification for finite impulse response systems under denial of service attacks

Optimal Stealthy Attack to Remote Estimator for Estimation Error Regulation.

DoS attack detection in identification of FIR systems with binary‐valued observations

False-Data-Injection-Enabled Network Parameter Modifications in Power Systems: Attack and Detection

Defense Optimization in Power Systems against False Data Injection Attacks

Tampering Attack Detection for Remote State Estimation With Quantized Observations

Fusion State Estimation of CPS Under Quantized Observations and Tampering Attacks

Tampering attack detection for remote interval observer

Event-Based Optimal Stealthy False Data-Injection Attacks Against Remote State Estimation Systems

Multi-Objective False Data Injection Attacks of Cyber–Physical Power Systems

Data-Based Tampered-Data Recovery Strategy With Encoding Against Stealthy Attack for Unknown Discrete-Time Systems

An Event-Based Stealthy Attack on Remote State Estimation

Improved Stealthy False Data Injection Attacks in Networked Control Systems

Optimal Attack Strategies Subject to Detection Constraints Against Cyber-Physical Systems

Event-Based Attack Against Remote State Estimation

Optimal innovation-based deception attacks with side information against remote state estimation in cyber-physical systems