Learning OTA: A Unified Framework for Edge Sensing, Computation, and Communication

2026-10-03

AI Summary

This study addresses the imbalance between resource efficiency and model accuracy in federated learning caused by the coupling of sensing, computation, and communication. We propose a unified over-the-air (OTA) federated learning framework that achieves the first joint modeling of these three processes. By deriving a non-convex convergence bound, we quantify the impact of multiple factors on the model trajectory. Furthermore, we formulate a mixed optimization strategy balancing energy consumption, latency, and learning quality, and design an efficient solver integrating alternating optimization, closed-form updates, and lightweight discrete search. Experimental results demonstrate that the proposed joint optimization scheme significantly enhances overall system performance and learning efficiency.

Abstract

Most federated learning (FL) studies implicitly assume that training data is readily available at participating edge devices. In practice, this data must first be acquired through heterogeneous sensors with potentially different measurement modalities, and the quality and cost of this acquisition process can directly affect the learning performance. This dependence creates a coupled relationship among sensing, computation, and communication, necessitating a system model that captures their joint effect on resource efficiency and learning accuracy. This paper develops such a unified treatment for an over-the-air (OTA) FL system in which each device configures its sensing modality, power, resolution, and sample size, with aggregations conducted OTA at the server. We derive models for sensing noise, energy consumption, latency, and OTA aggregation distortion, and establish a non-convex convergence bound quantifying how these factors influence the global model trajectory. Guided by this analysis, we formulate a per-round optimization that balances energy, latency, and learning quality under physical device constraints. The resulting mixed optimization is addressed through an alternating procedure with closed-form and one-dimensional updates for continuous variables and a lightweight discrete search. Numerical experiments corroborate the advantage of jointly optimizing sensing, computation, and communication decisions.

Problem

• Federated Learning
• Over-the-Air Computation
• Edge Sensing
• Resource Efficiency
• Joint Optimization

Innovation

• Federated Learning
• Over-the-Air Computation
• Joint Sensing-Communication-Computation
• Convergence Bound
• Mixed Optimization