Pinpointing Actuator Attacks: A Novel Diagnostic Framework for Cyber-Physical Systems

Zifan Wang, Qinru Qiu, Fanxin Kong

Research output: Chapter in Book/Entry/PoemConference contribution

Abstract

Attack defense is a critical research problem in Cyber-Physical System security. While significant research has focused on attack detection, the crucial aspect of attack diagnosis, particularly temporal diagnosis, remains underexplored. This paper addresses this critical gap by proposing a novel approach to actuator attack diagnosis. We present a real-time, data-driven solution for actuator attack diagnosis in CPS that operates without prior knowledge of the system and is not limited to specific attack types. Our approach leverages attention mechanisms and provides both temporal and spatial diagnosis. Through extensive testing using high-fidelity simulators and a case study on Raspberry Pi, we demonstrate the robustness, accuracy, and efficiency of our method. This research contributes to advancing CPS security and facilitating effective attack recovery strategies.

Original languageEnglish (US)
Title of host publicationProceedings - 2024 International Conference on Assured Autonomy, ICAA 2024
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages77-86
Number of pages10
ISBN (Electronic)9798331521011
DOIs
StatePublished - 2024
Event3rd International Conference on Assured Autonomy, ICAA 2024 - Nashville, United States
Duration: Oct 10 2024Oct 11 2024

Publication series

NameProceedings - 2024 International Conference on Assured Autonomy, ICAA 2024

Conference

Conference3rd International Conference on Assured Autonomy, ICAA 2024
Country/TerritoryUnited States
CityNashville
Period10/10/2410/11/24

Keywords

  • actuator attacks
  • cyber-physical systems
  • detection
  • diagnosis
  • real-time

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Decision Sciences (miscellaneous)
  • Safety, Risk, Reliability and Quality
  • Control and Optimization

Fingerprint

Dive into the research topics of 'Pinpointing Actuator Attacks: A Novel Diagnostic Framework for Cyber-Physical Systems'. Together they form a unique fingerprint.

Cite this