EUVIMEDEUVIMED entdecken
Crossref

Cyber-Physical Intrusion Detection System for Unmanned Aerial Vehicles

Samuel Chase Hassler, Umair Ahmad Mughal, Muhammad Ismail

IEEE Transactions on Intelligent Transportation Systems · 2024-06 · 25 (6) · 6106-6117

DOI: 10.1109/tits.2023.3339728 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Intelligent Transportation Systems, 2024-06. Verfasst von Samuel Chase Hassler und 2 weiteren Autor:innen. Die Arbeit wurde bisher 99-mal zitiert (Crossref) und nennt 31 Literaturangaben. Gefördert von: National Science Foundation (NSF) Energy, Power, Control, and Networks Program.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The increasing reliance on unmanned aerial vehicles (UAVs) has escalated the associated cyber risks. While machine learning has enabled intrusion detection systems (IDSs), current IDSs do not incorporate cyber-physical UAV features, which limits their detection performance. Additionally, the lack of public UAV’s cyber and physical datasets to develop IDS hinders further research. Therefore, this paper proposes a novel IDS fusing UAV cyber and physical features to improve detection capabilities. First, we developed a testbed that includes UAV, controller, and data collection tools to execute cyber-attacks and gather cyber and physical data under normal and attack conditions. We made this dataset publicly available. The dataset covers a range of cyber-attacks including denial-of-service, replay, evil twin, and false data injection attacks. Then, machine learning-based IDSs fusing cyber and physical features were trained to detect cyber-attacks using support vector machines, feedforward neural networks, recurrent neural networks with long short-term memory cells, and convolutional neural networks. Extensive experiments were conducted on varying complexity and range of attack training data to explore whether (a) fusion of cyber and physical features enhances detection performance compared to cyber or physical features alone, (b) fusion enhances detection when IDS is trained on a single attack type and tested on unseen attacks of varying complexity, (c) fusion enhances performance when the range of attack training data increases and models are tested on unseen attacks. Answering these research questions provides insights into IDS capabilities using cyber, physical, and cyber-physical features under different conditions.

Förderung

  • National Science Foundation (NSF) Energy, Power, Control, and Networks Program

Zitieren

APA

Hassler, S. C., Mughal, U. A., & Ismail, M. (2024). Cyber-Physical Intrusion Detection System for Unmanned Aerial Vehicles. IEEE Transactions on Intelligent Transportation Systems, 25(6), 6106-6117. https://doi.org/10.1109/tits.2023.3339728

Vancouver

Hassler SC, Mughal UA, Ismail M. Cyber-Physical Intrusion Detection System for Unmanned Aerial Vehicles. IEEE Transactions on Intelligent Transportation Systems. 2024;25(6):6106-6117. doi:10.1109/tits.2023.3339728

BibTeX
@article{Hassler2024,
  author = {Hassler, Samuel Chase and Mughal, Umair Ahmad and Ismail, Muhammad},
  title = {Cyber-Physical Intrusion Detection System for Unmanned Aerial Vehicles},
  journal = {IEEE Transactions on Intelligent Transportation Systems},
  year = {2024},
  volume = {25},
  number = {6},
  pages = {6106--6117},
  doi = {10.1109/tits.2023.3339728}
}
Kostenlos starten →