Federated learning over physical channels: adaptive algorithms with near-optimal guarantees

2025-09-02

AI Summary

To address high communication overhead and physical-layer transmission impairments—such as channel noise and hardware constraints—in federated learning, this paper proposes an adaptive Federated Stochastic Gradient Descent (FedSGD) algorithm tailored for Over-the-Air Computation (AirComp). The method introduces the first adaptive, noise-robust gradient aggregation scheme operating directly over noisy physical channels, dynamically adjusting the learning rate to compensate for channel-induced gradient distortion. We theoretically establish that the algorithm achieves a convergence rate asymptotically approaching the optimal bound of centralized SGD, while rigorously characterizing how convergence depends on both channel noise power and device heterogeneity (e.g., data and system non-IIDness). Extensive simulations with deep neural networks demonstrate the algorithm’s efficiency and robustness under realistic, non-ideal wireless channels and resource-constrained edge devices.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Search and Optimization: Learning to Search
• Planning, Routing, and Scheduling: Planning/Scheduling and Learning

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

In federated learning, communication cost can be significantly reduced by transmitting the information over the air through physical channels. In this paper, we propose a new class of adaptive federated stochastic gradient descent (SGD) algorithms that can be implemented over physical channels, taking into account both channel noise and hardware constraints. We establish theoretical guarantees for the proposed algorithms, demonstrating convergence rates that are adaptive to the stochastic gradient noise level. We also demonstrate the practical effectiveness of our algorithms through simulation studies with deep learning models.

Problem

• Develops adaptive federated SGD algorithms for physical channels
• Addresses channel noise and hardware constraints in federated learning
• Establishes convergence guarantees adaptive to gradient noise levels

Innovation

• Adaptive federated SGD over physical channels
• Algorithms handle channel noise and constraints
• Convergence rates adapt to gradient noise