Robust Federated Learning for Wireless Networks: A Demonstration with Channel Estimation

TL;DR

This paper addresses security vulnerabilities of federated learning in wireless channel estimation, proposing StoMedian as a robust aggregation method and LLPF as a data-level defense. StoMedian leverages a Bayesian-inspired stochastic filter with median-based weighting to mitigate adversarial updates, while LLPF uses a truncated Gaussian loss distribution to filter anomalous samples before training. Through simulations on a channel-estimation task, the authors show StoMedian and LLPF improve resilience against data-poisoning and various attack modes, preserving convergence where standard FedAvg fails. The work offers practical defenses for FL in dynamic wireless environments, with implications for secure, privacy-preserving channel estimation at scale.

Abstract

Federated learning (FL) offers a privacy-preserving collaborative approach for training models in wireless networks, with channel estimation emerging as a promising application. Despite extensive studies on FL-empowered channel estimation, the security concerns associated with FL require meticulous attention. In a scenario where small base stations (SBSs) serve as local models trained on cached data, and a macro base station (MBS) functions as the global model setting, an attacker can exploit the vulnerability of FL, launching attacks with various adversarial attacks or deployment tactics. In this paper, we analyze such vulnerabilities, corresponding solutions were brought forth, and validated through simulation.

Figures (4)

• Figure 1: CDF of actual samples and truncated Gaussian distribution
• Figure 2: Threat evaluation
• Figure 3: Convergence test of
• Figure 4: Convergence test of