Byzantine-Resilient Over-the-Air Federated Learning under Zero-Trust Architecture

TL;DR

This work tackles Byzantine resilience in over-the-air federated learning (AirFL) by embedding a zero-trust architecture (ZTA) and adaptive clustering into a robust federation scheme (FedSAC). It introduces a reputation-based Byzantine identification method, a sequential clustering strategy, and an adaptive weighting optimization solved via penalty convex-concave programming, all under Lyapunov drift to ensure long-term fairness. Theoretical one-step convergence analysis supports the design, and simulations on MNIST and CIFAR-10 demonstrate significant gains in accuracy and faster convergence compared with basinline approaches, even with few clusters. FedSAC thereby provides a practical, security-conscious framework for robust AirFL in 6G-enabled wireless networks.

Abstract

Over-the-air computation (AirComp) has emerged as an essential approach for enabling communication-efficient federated learning (FL) over wireless networks. Nonetheless, the inherent analog transmission mechanism in AirComp-based FL (AirFL) intensifies challenges posed by potential Byzantine attacks. In this paper, we propose a novel Byzantine-robust FL paradigm for over-the-air transmissions, referred to as federated learning with secure adaptive clustering (FedSAC). FedSAC aims to protect a portion of the devices from attacks through zero trust architecture (ZTA) based Byzantine identification and adaptive device clustering. By conducting a one-step convergence analysis, we theoretically characterize the convergence behavior with different device clustering mechanisms and uneven aggregation weighting factors for each device. Building upon our analytical results, we formulate a joint optimization problem for the clustering and weighting factors in each communication round. To facilitate the targeted optimization, we propose a dynamic Byzantine identification method using historical reputation based on ZTA. Furthermore, we introduce a sequential clustering method, transforming the joint optimization into a weighting optimization problem without sacrificing the optimality. To optimize the weighting, we capitalize on the penalty convex-concave procedure (P-CCP) to obtain a stationary solution. Numerical results substantiate the superiority of the proposed FedSAC over existing methods in terms of both test accuracy and convergence rate.

Figures (9)

• Figure 1: Architecture of existing robust aggregation methods via hierarchical AirComp and device clustering.
• Figure 2: Schematic diagram of the relationship between different device sets.
• Figure 3: Architecture of the proposed ZTA-empowered FedSAC approach.
• Figure 4: Test accuracy for the MNIST dataset and different types of attacks. (a) Sign-flipping attack (b) Gaussian attack (c) Label-flipping attack.
• Figure 5: Test accuracy for the CIFAR-10 dataset and different types of attacks. (a) Sign-flipping attack, (b) Gaussian attack, and (c) Label-flipping attack.