AirPASS: Over-the-Air Federated Learning via Pinching Antenna Systems

Jul 16, 2026

Upload by Sage

AirPASS investigates over-the-air federated learning (AirFL) for wireless networks where the access point uses a multi-waveguide pinching antenna system (PASS). The study adopts a learning-oriented AirFL goal: maximize the number of selected devices while keeping aggregation distortion under a target bound. Device selection, receive beamforming, and PASS placement form a highly nonconvex joint optimization problem. AirPASS proposes an alternating-optimization framework combining homotopy-Riemannian margin consolidation and homotopy-assisted geometric optimization, achieving strong results versus MIMO baselines and practical performance–complexity tradeoffs.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages13

Reading Time33 min

Introduction

Federated learning over wireless networks

Over-the-air federated learning and AirComp

Related work in AirFL system design

Learning-oriented AirFL formulation

Pinching antenna systems (PASS) background

AirPASS: Over-the-Air Federated Learning via Pinching Antenna Systems Seyed Mohammad Azimi-Abarghouyi, Member, IEEE and Christopher G. Brinton, Senior Member, IEEE arXiv :2607 .06768v 1 [ cs .IT] 7 Jul 2026 Abstract—This paper investigates over-the-air federated learning (AirFL) in wireless systems where the access point is equipped with a multi-waveguide pinching antenna system (PASS). We adopt the widely studied learning-oriented AirFL formulation, which seeks to maximize the number of selected devices while keeping the aggregation distortion below a prescribed threshold. The resulting joint optimization of device selection, receive beamforming, and pinching-antenna placement is highly nonconvex due to the intricate coupling among these system variables. To address this challenge, we develop AirPASS, an alternating optimization framework with two main components: a homotopyRiemannian margin-consolidation method for device selection and receive beamforming under ﬁxed PASS conﬁguration, anda homotopy-assisted geometry optimization method for updating the pinching-antenna positions under ﬁxed selected devices and beamformer. Experiments show that AirPASS consistently outperforms conventional co-located MIMO baselines, remains close to ideal FedAvg, and achieves an attractive performance– complexity tradeoff relative to SDR-DC and matching-pursuit scheduling alternatives. Index Terms—Over-the-air federated learning, over-the-air computation, pinching antenna systems, device selection, receive beamforming, Riemannian optimization. I. INTRODUCTION Federated learning (FL) enables multiple devices to collaboratively train a shared machine learning model while keeping their raw data local [1] . Devices perform local updates using private datasets and periodically transmit model updates to a central server for aggregation into a global model. This paradigm is particularly attractive for wireless networks, where data is generated at the edge and privacy concerns often limit centralized data collection. However, deploying FL over wireless networks introduces fundamental communication challenges. Modern learning models typically contain millions of parameters, and exchanging model updates over many training rounds can incur signiﬁcant communication latency and bandwidth overhead. These limitations motivate the design of communication-efﬁcient learning mechanisms that tightly integrate wireless communication and distributed training. A promising solution is over-the-air federated learning (AirFL), which builds upon the principle of over-the-air computation (AirComp) [2] . AirComp exploits the waveform superposition property of wireless multiple-access channels to compute functions of distributed signals directly in the air S. M. Azimi-Abarghouyi is with the Department of Electrical Engineering, Chalmers University of Technology, Gothenburg, Sweden (Email: [azimimo@chalmers.se](azimimo@chalmers.se)) . C. G. Brinton is with the School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN USA (Email: [cgb@purdue.edu](cgb@purdue.edu)). [3] . Recently, alternative perspectives on exploiting wireless superposition have also begun to emerge, including out-ofair computation (AirCPU) [4] . In AirFL systems, devices simultaneously transmit analog-modulated model updates, and the access point (AP) directly receives their superposition, which corresponds to the aggregated model update required for FL. By integrating communication and aggregation atthe physical layer, AirFL signiﬁcantly reduces communication latency compared to conventional orthogonal multiple access aggregation schemes. Recent years have witnessed extensive research on AirFL system design. Early works studied FL over wireless fading channels and demonstrated the advantages of analog aggregation [5] . Subsequent studies investigated AirFL designs that jointly optimize wireless transmission and learning performance. In particular, receive bea

FAQs

What optimization objective does AirPASS target in over-the-air federated learning?

AirPASS uses a learning-oriented formulation that maximizes the number of selected devices while constraining the AirComp aggregation distortion to stay below a prescribed threshold.

Why is the joint design in AirPASS considered difficult?

Device selection, receive beamforming, and pinching-antenna placement are coupled, making the overall problem highly nonconvex and challenging to optimize directly.

How does AirPASS solve the nonconvex problem?

AirPASS employs alternating optimization with two core stages: homotopy-Riemannian margin consolidation for device selection and beamforming under fixed PASS, and homotopy-assisted geometry optimization for updating antenna positions under fixed devices and beamformers.

Jul 16, 2026

Upload by Sage