Skip to Main Content
 

Global Search Box

 
 
 

ETD Abstract Container

Abstract Header

Cybersecurity for Networked Control System

Tang, Jiacheng
http://rave.ohiolink.edu/etdc/view?acc_num=osu1649254687505393

Abstract Details

2022, Doctor of Philosophy, Ohio State University, Electrical and Computer Engineering.
This dissertation studies cybersecurity, with a focus on the control theory for cyber physical system (CPS). We consider CPS as a networked control system (NCS) where the sensing and control signals transmitted through the communication network can be compromised. To secure the NCS, we develop detection and response algorithms against adversarial attacks. In particular, we propose dynamic watermarking as an active detection scheme against sensor spoofing attacks, stochastic transmission of the control signals as a control response against bus-off (DoS) attack, and optimization on communication link elimination as a control response against general attack scenarios whenever network switching is needed. In all the cases, we show the efficacy of applying the proposed algorithms against attacks and the guarantee of control performance after applying security measures. For communication link elimination, we consider a group of agent switching from a fully connected communication network to a backup network with limited bandwidth due to attack. This problem is formulated as a linear quadratic Gaussian (LQG) team problem. We show that optimizing communication links in the backup network is equivalent to adding a cardianlity constraint to the original quadratic optimization problem. Due to the combinatorial nature of the problem, we apply convex relaxation on the constraint set and random projection on the objective function as dimenaionality reduction tools. We provide the theoretical performance guarantee on the near-optimal solution obtained from the relaxed problem. For stochastic transmission against bus-off attacks, we present a mathematical model for the bus-off attack and formulate it as a non-zero sum game between the controller-transmitter pair and the attacker. A stochastic transmission policy is proposed as a proactive countermeasure when the attacker persists in the network. We determine the Nash equilibria of the non-zero sum game in the cases off open-loop and closed-loop attacker. To demonstrate the effectiveness of our theoretical results, we apply the stochastic transmission policy on an adaptive cruise control (ACC) and show that under the Nash equilibrium strategies, an appropriate stochastic transmission policy stabilizes the error counter and the ego vehicle maintains a safe distance with the leading vehicle. For dynamic watermarking against spoofing attacks, we extend the exisiting dynamic watermarking algorithm to a finite Markov decision process (MDP) setup. The watermark is added by changing the control policy of the controller. Compared with the linear/nonlinear model with dynamic watermarking, MDP model allows for more abstract elements (e.g., active and sleep) in the state and action space. Here, we propose a CUSUM-type detection scheme to detect whether the information coming from the feedback channel is authentic or has been compromised. We analyze the performance bound of the detection algorithm in terms of the detection delay and false alarms, and also the control loss due to introducing such watermarks into the system. We apply dynamic watermarking algorithm to detect a spoofing attack on a power management system in a sensor network. We demonstrate the trade-off between detection performance and control loss for different magnitude of the watermark applied to the system.
Abhishek Gupta (Advisor)
Emre Koksal (Committee Member)
Aylin Yener (Committee Member)
111 p.
Electrical Engineering
Cybersecurity, Control System

Recommended Citations

Citations

  • Tang, J. (2022). Cybersecurity for Networked Control System [Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1649254687505393

    APA Style (7th edition)

  • Tang, Jiacheng. Cybersecurity for Networked Control System. 2022. Ohio State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1649254687505393.

    MLA Style (8th edition)

  • Tang, Jiacheng. "Cybersecurity for Networked Control System." Doctoral dissertation, Ohio State University, 2022. http://rave.ohiolink.edu/etdc/view?acc_num=osu1649254687505393

    Chicago Manual of Style (17th edition)

osu1649254687505393
149
© 2022, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.