TASER: Task-Aware Spectral Energy Refine for Backdoor Suppression in UAV Swarms Decentralized Federated Learning
Sizhe Huang, Shujie Yang

TL;DR
TASER is a decentralized spectral analysis method that effectively detects and suppresses stealthy backdoor attacks in UAV federated learning by focusing on spectral concentration differences, outperforming traditional outlier-based defenses.
Contribution
This paper introduces TASER, the first spectral concentration-based decentralized defense for backdoor suppression in UAV federated learning, addressing limitations of outlier detection methods.
Findings
TASER reduces attack success rate below 20%
Achieves less than 5% accuracy loss on benign tasks
Effective against stealthy backdoor attacks bypassing traditional defenses
Abstract
As backdoor attacks in UAV-based decentralized federated learning (DFL) grow increasingly stealthy and sophisticated, existing defenses have likewise escalated in complexity. Yet these defenses, which rely heavily on outlier detection, remain vulnerable to carefully crafted backdoors. In UAV-DFL, the lack of global coordination and limited resources further render outlier-based defenses impractical. Against this backdrop, gradient spectral analysis offers a promising alternative. While prior work primarily leverages low-frequency coefficients for pairwise comparisons, it neglects to analyze the intrinsic spectral characteristics of backdoor gradients. Through empirical analysis of existing stealthy attacks, we reveal a key insight: the more effort attackers invest in mimicking benign behaviors, the more distinct the spectral concentration becomes. Motivated by this, we propose…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdversarial Robustness in Machine Learning · UAV Applications and Optimization · Privacy-Preserving Technologies in Data
