TASER: Task-Aware Spectral Energy Refine for Backdoor Suppression in UAV Swarms Decentralized Federated Learning

TL;DR

This work proposes Task-Aware Spectral Energy Refine (TASER), the first efficient backdoor defense that utilizes spectral concentration instead of complex outlier detection, enabling mitigation of stealthy attacks by structurally disrupting the backdoor task.

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 Task-Aware Spectral Energy Refine (TASER) -- a decentralized defense framework. To our knowledge, this is the first efficient backdoor defense that utilizes spectral concentration instead of complex outlier detection, enabling mitigation of stealthy attacks by structurally disrupting the backdoor task. To suppress the backdoor task, TASER preserves main-task-relevant frequency coefficients and discards others. We provide theoretical guarantees and demonstrate through experiments that TASER remains effective against stealthy backdoor attacks that bypass outlier-based defenses, achieving attack success rate below 20% and accuracy loss under 5%.

Figures (7)

• Figure 1: High-energy frequency gradient differences between traditional and stealthy backdoor attacks. We tracked the high-energy frequency indices of 50 clean clients and 50 malicious clients over 400 communication rounds and aggregated their selections into a frequency--time heatmap. The x-axis represents the frequency index, and the y-axis represents the communication round. Blue regions indicate frequencies more frequently selected as high-energy by malicious clients, while red regions indicate those favored by benign clients.
• Figure 2: Workflow of TASER
• Figure 3: Comparison of main task accuracy and backdoor success rate under different frequency selection ratios.
• Figure 4: CIFAR-10 against black-box stealthy backdoor attacks.
• Figure 5: CIFAR-10 against white-box stealthy backdoor attacks.