Pinching Antennas-Assisted Low-Latency Federated Learning Over Multi-User Wireless Networks

2026-03-09

AI Summary

This work addresses the challenge of unreliable communication in wireless federated learning, which often arises from uplink congestion, channel fading, or poor line-of-sight conditions, leading to high training latency and degraded model performance. To mitigate these issues, the authors propose FedPASS, a novel framework that integrates a Pinching antenna system—capable of dynamically adjusting its radiation point—into federated learning for the first time. FedPASS jointly optimizes device scheduling, transmit power, computation frequency, and antenna positioning to enhance line-of-sight links and collaboratively minimize end-to-end training delay. The resulting mixed-integer nonlinear programming problem is efficiently solved via a combination of block coordinate descent and Gauss–Seidel grid search algorithms. Experiments demonstrate that FedPASS achieves accuracy comparable to an ideal baseline on MNIST and CIFAR-10 while significantly reducing total training latency.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Federated learning (FL) over wireless networks is fundamentally constrained by unreliable communication links, particularly when uplink channels suffer from blockage, fading, or weak line-of-sight (LoS) conditions. Pinching-antenna systems (PASSs) offer a new physical-layer capability to dynamically reposition radiating points along a dielectric waveguide, enabling controllable LoS connectivity and significantly improved channel quality. This paper develops FedPASS, a novel framework for low-latency wireless FL assisted by PASS. We formulate a multi-objective optimization problem that jointly minimizes the end-to-end round latency and an upper bound on the FL optimality gap. The resulting formulation is a mixed-integer nonlinear program subject to practical constraints on scheduling, transmit power, local CPU frequency, and PA placement. To address the resulting computational challenges, we develop a two-tier iterative algorithm: an outer loop that updates scheduling, communication time allocation, and power control via block coordinate descent, and an inner loop that optimizes PA locations using a Gauss-Seidel-based coordinate update with grid search under spacing constraints. Numerical results on MNIST and CIFAR-10 demonstrate that FedPASS achieves accuracy comparable to idealized FL baselines while drastically reducing the total training latency compared to conventional wireless FL.

Problem

• federated learning
• wireless networks
• low-latency
• unreliable communication
• line-of-sight

Innovation

• Pinching Antenna Systems
• Federated Learning
• Low-Latency Wireless Communication
• Multi-Objective Optimization
• Block Coordinate Descent