Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Data-driven Supervisory Control under Attacks via Spectral Learning

Nathaniel Smith, Yu Wang

TL;DR

This work tackles the challenge of securing CPS against unknown, broad-spectrum attacks by modeling the plant, supervisor, and attackers as finite-state transducers. It introduces a data-driven, spectral learning method to infer attacker FSTs from observed attack histories, enabling the synthesis of a resilient supervisor that counteracts sensor and actuator attacks. The core contribution is a learning framework using Hankel-like matrices and basis masks to recover the attacker dynamics, ensuring the learned models faithfully reproduce the attack language. The approach aims to deliver robust supervisory control that remains effective under unseen attack patterns, with practical relevance for cyber-physical security in diverse CPS domains.

Abstract

The technological advancements facilitating the rapid development of cyber-physical systems (CPS) also render such systems vulnerable to cyber attacks with devastating effects. Supervisory control is a commonly used control method to neutralize attacks on CPS. The supervisor strives to confine the (symbolic) paths of the system to a desired language via sensors and actuators in a closed control loop, even when attackers can manipulate the symbols received by the sensors and actuators. Currently, supervisory control methods face limitations when effectively identifying and mitigating unknown, broad-spectrum attackers. In order to capture the behavior of broad-spectrum attacks on both sensing and actuation channels we model the plant, supervisors, and attackers with finite-state transducers (FSTs). Our general method for addressing unknown attackers involves constructing FST models of the attackers from spectral analysis of their input and output symbol sequences recorded from a history of attack behaviors observed in a supervisory control loop. To construct these FST models, we devise a novel learning method based on the recorded history of attack behaviors. A supervisor is synthesized using such models to neutralize the attacks.

Data-driven Supervisory Control under Attacks via Spectral Learning

TL;DR

This work tackles the challenge of securing CPS against unknown, broad-spectrum attacks by modeling the plant, supervisor, and attackers as finite-state transducers. It introduces a data-driven, spectral learning method to infer attacker FSTs from observed attack histories, enabling the synthesis of a resilient supervisor that counteracts sensor and actuator attacks. The core contribution is a learning framework using Hankel-like matrices and basis masks to recover the attacker dynamics, ensuring the learned models faithfully reproduce the attack language. The approach aims to deliver robust supervisory control that remains effective under unseen attack patterns, with practical relevance for cyber-physical security in diverse CPS domains.

Abstract

The technological advancements facilitating the rapid development of cyber-physical systems (CPS) also render such systems vulnerable to cyber attacks with devastating effects. Supervisory control is a commonly used control method to neutralize attacks on CPS. The supervisor strives to confine the (symbolic) paths of the system to a desired language via sensors and actuators in a closed control loop, even when attackers can manipulate the symbols received by the sensors and actuators. Currently, supervisory control methods face limitations when effectively identifying and mitigating unknown, broad-spectrum attackers. In order to capture the behavior of broad-spectrum attacks on both sensing and actuation channels we model the plant, supervisors, and attackers with finite-state transducers (FSTs). Our general method for addressing unknown attackers involves constructing FST models of the attackers from spectral analysis of their input and output symbol sequences recorded from a history of attack behaviors observed in a supervisory control loop. To construct these FST models, we devise a novel learning method based on the recorded history of attack behaviors. A supervisor is synthesized using such models to neutralize the attacks.

Paper Structure

This paper contains 6 sections, 4 equations, 4 figures.

Table of Contents

  1. Introduction
  2. Preliminaries
  3. Finite state transducers
  4. Supervisory control with FSTs
  5. Problem Formulation
  6. Data Requirements

Figures (4)

  • Figure 1: Supervisory control of cyber-physical systems under attacks. Attackers intercept messages between Plant and Supervisor. Attacks recorded during testing phases.
  • Figure 2: An Example FST.
  • Figure 3: Supervisory control loop under sensor and actuator attacks. Sensor attacker manipulates messages from plant to supervisor. Actuator attacker manipulates messages from supervisor to plant
  • Figure 4: Example of supervisory control loop with an actuator attacker $\mathcal{A}_a$.

Theorems & Definitions (5)

  • Definition 1
  • Example 1
  • Example 2
  • Definition 2
  • Example 3