Yuzhe Li

Jamming Attacks on Remote State Estimation in Cyber-Physical Systems: A Game-Theoretic Approach

We consider security issues in remote state estimation of Cyber-Physical Systems (CPS). A sensor node communicates with a remote estimator through a wireless channel which may be jammed by an external attacker. With energy constraints for both the sensor and the attacker, the interactive decision making process of when to send and when to attack is studied. We formulate a game-theoretic framework and prove that the optimal strategies for both sides constitute a Nash equilibrium of a zero-sum game. To tackle the computation complexity issues, we present a constraint-relaxed problem and provide corresponding solutions using Markov chain theory.

Jamming attack on Cyber-Physical Systems: A game-theoretic approach

We consider security issues in Cyber-Physical Systems (CPSs). A sensor node communicates with a remote estimator through a wireless channel which may be jammed by an external attacker. With energy constraints for both the sensor and the attacker, the interactive decision making process of when to send and when to attack is studied. We formulate a game-theoretical framework and prove that the optimal strategies for both sides constitute a Nash equilibrium. The derivation of the optimal strategies for both sides is also provided with examples.

Optimal Periodic Transmission Power Schedules for Remote Estimation of ARMA Processes

We consider periodic sensor transmission power allocation with an average energy constraint. The sensor sends its Kalman filter-based state estimate to the remote estimator through an unreliable link. Dropout probabilities depend on the power level used. To encompass applications where the estimator needs to attend to multiple tasks, we allow for irregular sampling, following a periodic pattern. Using properties of an underlying Markov chain model, we derive an explicit expression for the estimation error covariance. The results are then used to study optimal sensor power scheduling which minimizes the average error covariance.

SINR-Based DoS Attack on Remote State Estimation: A Game-Theoretic Approach

We consider remote state estimation of cyberphysical systems under signal-to-interference-plus-noise ratio-based denial-of-service attacks. A sensor sends its local estimate to a remote estimator through a wireless network that may suffer interference from an attacker. Both the sensor and the attacker have energy constraints. We first study an associated two-player game when multiple power levels are available. Then, we build a Markov game framework to model the interactive decision-making process based on the current state and information collected from previous time steps. To solve the associated optimality (Bellman) equations, a modified Nash Q-learning algorithm is applied to obtain the optimal solutions. Numerical examples and simulations are provided to demonstrate our results.

Data-driven power control for state estimation

We consider sensor transmission power control for state estimation, using a Bayesian inference approach. A sensor node sends its local state estimate to a remote estimator over an unreliable wireless communication channel with random data packet drops. As related to packet dropout rate, transmission power is chosen by the sensor based on the relative importance of the local state estimate. The proposed power controller is proved to preserve Gaussianity of local estimate innovation, which enables us to obtain a closed-form solution of the expected state estimation error covariance. Comparisons with alternative non-data-driven controllers demonstrate performance improvement using our approach.

Online sensor transmission power schedule for remote state estimation

We propose an online sensor transmission power schedule for remote state estimation. A sensor sends its local state estimate to a remote estimator through an unreliable wireless channel, which introduces random data packet drops. The packet dropout rate is related to the transmission power which is allocated by the sensor under an energy constraint. The sensor chooses the transmission power based on the relative importance of the local estimate at each time. We prove that the proposed power schedule preserves the Gaussian distribution of the local estimate innovation, which enable us to obtain a closed-form solution of the expected state estimation error covariance. Comparisons with alternative offline schedules are provided, which demonstrate significant performance improvement by the online schedule.

A multi-channel transmission schedule for remote state estimation under DoS attacks

This paper considers a cyber-physical system (CPS) under denial-of-service (DoS) attacks. The measurements of a sensor are transmitted to a remote estimator over a multi-channel network, which may be congested by a malicious attacker. Among these multiple communication paths with different characteristics and properties at each time step, the sensor needs to choose a single channel for sending data packets while reducing the probability of being attacked. In the meanwhile, the attacker needs to decide the target channel to jam under an energy budget constraint. To model this interactive decision-making process between the two sides, we formulate a two-player zero-sum stochastic game framework. A Nash Q-learning algorithm is proposed to tackle the computation complexity when solving the optimal strategies for both players. Numerical examples are provided to illustrate the obtained results.

Detection Against Linear Deception Attacks on Multi-Sensor Remote State Estimation

In this paper, a security problem in cyberphysical systems (CPS) is studied. A remote state estimation process using multiple sensors is considered. The measurement innovation packets from each sensor, which may be modified by a malicious attacker, are sent to a remote fusion center through wireless communication channels. To avoid being detected by typical bad data detectors at the remote estimators side, the attacker would maintain the statistical properties of the measurements. Based on the information extracted from the trusted sensors and the correlations between the trusted sensors and the suspicious sensors, we propose three sequential data verification and fusion procedures for different detection information scenarios. The corresponding impacts of possible attacking patterns on the estimation performance under different detectors are analyzed explicitly. Simulations are provided to illustrate the developed results.

A Game-Theoretic Approach to Fake-Acknowledgment Attack on Cyber-Physical Systems

A class of malicious attacks against remote state estimation in cyber-physical systems is considered. A sensor adopts an acknowledgement (ACK)-based online power schedule to improve the remote state estimation performance under limited resources. To launch malicious attacks, the attacker can modify the ACKs from the remote estimator and convey fake information to the sensor, thereby misleading the sensor with subsequent performance degradation. One feasible attack pattern is proposed and the corresponding effect on the estimation performance is derived analytically. Due to the ACKs being unreliable, the sensor needs to decide at each instant, whether to trust the ACK information or not and adapt the transmission schedule accordingly. In the meanwhile, there is also a tradeoff for the attacker between attacking and not attacking when the modification of ACKs is costly. To investigate the optimal strategies for both the sensor and the attacker, a game-theoretic framework is built and the equilibrium for both sides is studied.

Power Control of an Energy Harvesting Sensor for Remote State Estimation

We investigate sensor transmission power control for remote state estimation. Instead of using a conventional sensor, a sensor equipped with an energy harvester which can obtain energy from the external environment is utilized. We formulate power control of the energy harvesting sensor into an infinite time-horizon Markov decision process (MDP). To deal with the computation complexity associated with this multi-dimensional MDP, a continuous-time approach and perturbation analysis are used and a closed-form approximate value function is derived. Based on the approximation, we obtain a closed-form optimal power control solution which has a threshold-based structure. A numerical example is provided to evaluate the estimation performance of the optimal solution compared with other power scheduling schemes.