BLAST: A Stealthy Backdoor Leverage Attack Against Cooperative Multi-Agent Deep Reinforcement Learning-Based Systems
IEEE Transactions on Cognitive Communications and Networking · 2026 · 12 · 9312-9325
DOI: 10.1109/tccn.2026.3706546 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Cognitive Communications and Networking, 2026. Verfasst von Jing Fang und 5 weiteren Autor:innen. Die Arbeit nennt 45 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Guangdong Basic and Applied Basic Research Foundation, talents of Nanjing University of Aeronautics and Astronautics.
Abstract (Originalsprache)
Recent studies have shown that cooperative multi-agent deep reinforcement learning (c-MADRL) is under the threat of backdoor attacks. Once a backdoor trigger is observed, it will perform malicious actions, resulting in failures or achieving malicious goals. However, existing backdoor attacks suffer from several issues, e.g., instant trigger patterns lack stealthiness, the backdoor is trained or activated by an additional network, or all agents are backdoored. To this end, in this paper, we propose a novel $\boldsymbol {B}$ ackdoor $\boldsymbol {L}$ everage $\boldsymbol {A}$ ttack againST c-MADRL, BLAST, which attacks the entire multi-agent team by embedding the backdoor only in a single agent. Firstly, we introduce adversary spatiotemporal behavior patterns as the backdoor trigger rather than manually injected fixed visual patterns or instant status and control the period to perform malicious actions. This method can guarantee the stealthiness and practicality of BLAST. Secondly, we hack the original reward function of the backdoor agent via unilateral guidance to inject BLAST, to achieve the leverage attack effect that can pry open the entire multi-agent system via a single backdoor agent. We evaluate our BLAST against 3 classic c-MADRL algorithms (VDN, QMIX, and MAPPO) in 2 popular c-MADRL environments (SMAC and Pursuit), and 3 existing defense mechanisms. The experimental results demonstrate that BLAST can achieve a high attack success rate while maintaining a low clean performance variance rate.
Förderung
- National Natural Science Foundation of China
- Guangdong Basic and Applied Basic Research Foundation
- talents of Nanjing University of Aeronautics and Astronautics
Zitieren
APA
Fang, J., Yan, S., Yin, X., Yu, Y., Tian, C., & Liu, J. (2026). BLAST: A Stealthy Backdoor Leverage Attack Against Cooperative Multi-Agent Deep Reinforcement Learning-Based Systems. IEEE Transactions on Cognitive Communications and Networking, 12, 9312-9325. https://doi.org/10.1109/tccn.2026.3706546
Vancouver
Fang J, Yan S, Yin X, Yu Y, Tian C, Liu J. BLAST: A Stealthy Backdoor Leverage Attack Against Cooperative Multi-Agent Deep Reinforcement Learning-Based Systems. IEEE Transactions on Cognitive Communications and Networking. 2026;12:9312-9325. doi:10.1109/tccn.2026.3706546
BibTeX
@article{Fang2026,
author = {Fang, Jing and Yan, Saihao and Yin, Xueyu and Yu, Yinbo and Tian, Chunwei and Liu, Jiajia},
title = {BLAST: A Stealthy Backdoor Leverage Attack Against Cooperative Multi-Agent Deep Reinforcement Learning-Based Systems},
journal = {IEEE Transactions on Cognitive Communications and Networking},
year = {2026},
volume = {12},
pages = {9312--9325},
doi = {10.1109/tccn.2026.3706546}
}
EUVIMED