Analog Over-the-Air Federated Learning with Interference-Based Energy Harvesting

2025-09-12

AI Summary

In over-the-air federated learning (OTA-FL) simulations, strong co-channel interference (CCI) induced by in-band radio-frequency energy harvesting causes significant aggregation errors and degrades learning performance. To address this, we propose a channel-state-information (CSI)-free energy-efficiency optimization framework. Our contributions are threefold: (1) an interference-aware, CSI-free denoising strategy that substantially reduces CCI-induced aggregation bias; (2) an adaptive device scheduling algorithm jointly leveraging device energy states and channel interference levels to dynamically optimize local training rounds and participation; and (3) an integrated mechanism coordinating energy harvesting, analog OTA computation, and convergence guarantees. Simulation results demonstrate that the proposed denoising method achieves performance close to CSI-dependent schemes; the adaptive scheduler accelerates model convergence by 32%; and increasing the number of active devices effectively mitigates high-power CCI—thereby preserving learning accuracy while reducing system energy consumption.

Technology Category

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

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

We consider analog over-the-air federated learning, where devices harvest energy from in-band and out-band radio frequency signals, with the former also causing co-channel interference (CCI). To mitigate the aggregation error, we propose an effective denoising policy that does not require channel state information (CSI). We also propose an adaptive scheduling algorithm that dynamically adjusts the number of local training epochs based on available energy, enhancing device participation and learning performance while reducing energy consumption. Simulation results and convergence analysis confirm the robust performance of the algorithm compared to conventional methods. It is shown that the performance of the proposed denoising method is comparable to that of conventional CSI-based methods. It is observed that high-power CCI severely degrades the learning performance, which can be mitigated by increasing the number of active devices, achievable via the adaptive algorithm.

Problem

• Mitigates co-channel interference in analog federated learning
• Proposes denoising policy without requiring channel state information
• Develops adaptive scheduling algorithm for energy-efficient device participation

Innovation

• Denoising policy without requiring CSI
• Adaptive scheduling algorithm for energy
• Mitigates interference via device participation