Low-Altitude Fluid Antenna Network with Multi-Agent Reinforcement Learning

Tong Zhang; Yanfei Su; Shuai Wang; Wanli Ni; Chengzhong Xu; Huseyin Arslan

cs.ITAug 28, 2026

Low-Altitude Fluid Antenna Network with Multi-Agent Reinforcement Learning

Authors: Tong Zhang, Yanfei Su, Shuai Wang, Wanli Ni, Chengzhong Xu, Huseyin Arslan

Organizations: Guangdong Provincial Key Laboratory of Aerospace Communication and Networking Technology, Harbin Institute of Technology, Shenzhen, 518055, China · Harbin Institute of Technology, Shenzhen, China · Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China · School of Information and Communication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China ·

Abstract

Low-altitude wireless networks (LAWNs) integrate terrestrial and aerial platforms to provide ubiquitous communication, sensing, and localization services for unmanned aerial vehicles (UAVs) and electric vertical takeoff and landing (eVTOL) aircraft. However, dynamic air-ground and air-air channels, abrupt blockages, and heterogeneous interference hinder the realization of this goal. Nevertheless, fluid antenna (FA), a cutting-edge multiple-input multiple-output (MIMO) technique, overcomes these challenges by reconfiguring antenna positions to unlock additional spatial degrees-of-freedom. In this paper, towards bringing low-altitude FA networks into reality, we study the fast and high-performance FA reconfiguration for low-altitude FA networks with multi-agent reinforcement learning (MARL). Specifically, we present an electromagnetic digital twin (EM-DT)-assisted MARL framework. To fill the sim-to-real gap, we introduce a two-stage transfer learning framework. Our case study shows that joint FA positions and beamforming optimization can enhance the system sum-rate by 118.5%, compared to the fixed position baseline. This gain comes from the dynamic millisecond timescale reconfiguration of FA arrays and the adaptive steering of beams toward aerial users with mobility.

Figures & tables

Panel of Figure 1 — Fig. 1 : Illustration of a low-altitude FA network (Subfig. a ), and conceptual comparison between traditional MIMO and FA MIMO (Subfigs. 1(c) and 1(c) ).

Panel of Figure 2 — Fig. 2 : Representative EM-DTs and their satellite imagery. The top row shows the satellite imagery and the bottom row shows the corresponding EM-DTs created by RANPLAN ACADEMIC V7.1 software, with Shenzhen University Town, China (Subfigs. 2(a) and 2(e) ), Guangzhou Zhujiang New Town CBD, China (Subfigs. 2(b) and 2(f) ), Wembley Stadium in London, UK (Subfigs. 2(c) and 2(g) ), and Las Vegas Strip, USA (Subfigs. 2(d) and 2(h) ).

Fig. 3 : Illustration of MARL training with the EM-DT, and typical Dec-POMDP component definitions.

Panel of Figure 4 — Fig. 4 : Case study of low-altitude FA networks with EM-DT-assisted MARL. Using FAs achieves $118.5\%$ network sum-rate gain under a mobility and unseen environment during testing.

	FA Reconfiguration Overhead	CSI Usage Reduction	Trustworthy MARL	EM-DT Fidelity	Heterogeneous Time-Scale	System Support for Artificial Intelligence
Open Issues	Reconfiguration cost & budget	CSI overhead CSI aging	Black-box policy Certification barriers	Static maps Dynamic blockages	Different dynamics Time-scale mismatch	Heterogeneous data values
Future Directions	Predictive positioning & hardware co-design	CSI prediction by EM-DT	Safe MARL Certification testbeds	Surrogate models Online calibration	Hierarchical MARL Asynchronous training	Task-oriented end-to-end design

TABLE I : Summary of open issues and future directions.

Fig. 5 : Low-altitude FA networks for aerial visual question answering.

Tong Zhang [M] ([email protected]) received the B.S. degree from Northwest University, Xi’an, China, in 2012, the M.S. degree from Beijing University of Posts and Telecommunications, Beijing, China, in 2015, and the Ph.D. degree in electronic engineering from The Chinese University of Hong Kong, Hong Kong, in 2020. He was a Post-Doctoral Fellow with the Southern University of Science and Technology from 2020 to 2022. He was a Lecturer with the Department of Electronic Engineering, Jinan University. He is currently an Assistant Professor with Harbin Institute of Technology, Shenzhen. His research interests include fluid antenna systems and reinforcement learning.

Table 7

Yanfei Su ([email protected]) received the B.S. degree in communication engineering from Harbin Institute of Technology (HIT), Weihai, China. She is currently pursuing the M.S. degree in information and communication engineering with HIT, Shenzhen, China. Her research interests include fluid antenna systems and reinforcement learning.

Table 8

Shuai Wang [M] ([email protected]) received the B.Eng. and M. Eng. degrees from Beijing University of Posts and Telecommunications ( BUPT), in 2011 and 2014 respectively, and the Ph.D. degree from University of Hong Kong (HKU) in 2018. He is now a Professor with the Shenzhen Institutes of Advanced Technology (SIAT), Chinese Academy of Sciences, where he leads the Intelligent Networked Vehicle Systems (INVS) Laboratory. His research intrests include robot learning and networked intelligence.

Table 9

Wanli Ni [M] ([email protected]) is currently an Assistant Professor with the School of Information and Communication Engineering, Beijing University of Posts and Telecommunications (BUPT), China. He received the B.Eng. degree in 2018 and the Ph.D. degree in 2023, both from BUPT. From 2022 to 2023, he was a visiting student at the Nanyang Technological University, Singapore. From 2023 to 2025, he was a Postdoctoral Researcher at the Tsinghua University, China. His research interests include federated learning, large AI models, spectrum sensing, and satellite communications.

Table 10

Chengzhong Xu [F] ([email protected]) received the Ph.D. degree from the University of Hong Kong, Hong Kong, in 1993. He has held Faculty positions with Wayne State University, Detroit, MI, USA, and the Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China. He is currently the Chair Professor of computer science with the University of Macau, Macau, China. He is also the Dean of Faculty of Information Science and Computing and the Director of Institute of AI and Brain Sciences. His research interests include cloud and edge for AI, autonomous driving, and intelligent transportation.

Table 11

Hüseyin Arslan [F] ([email protected]) is currently the Dean of the School of Engineering and Natural Sciences at Istanbul Medipol University, Türkiye. He received his bachelor’s degree from Middle East Technical University, Türkiye, in 1992, and his M.S. and Ph.D. degrees from Southern Methodist University, USA, in 1994 and 1998, respectively. From 1998 to 2002, he was with Ericsson Research, where he worked on 2G and 3G wireless communication systems. From 2002 to 2022, he was with the University of South Florida, USA, where he served as a Professor of Electrical Engineering. In 2013, he joined Istanbul Medipol University and contributed to the establishment of its engineering school. He has also held various academic, industrial, and advisory roles and has served on the editorial boards of several IEEE journals. His research interests include 6G and beyond wireless communications, waveform design, spectrum sharing, physical-layer security, interference management, non-terrestrial networks, and integrated sensing and communication.

Table 12

Wireless Communications

Wireless communications studies how information is sent over radio links and how networks manage signals, spectrum, and resources reliably.

See all 391 papers

Agentic AI Networking for Heterogeneous Unmanned Aerial Systems in Low-Altitude Wireless Networks
Sep 17, 2026Nguyen Duc Minh Quang, Chang Liu, Shuangyang Li +1LLM-Guided RL Wireless Communications
School of Computing, Engineering, and Mathematical Sciences, La Trobe University, Melbourne, VIC, Australia · Chair of Communications and Information Theory, Technical University of Berlin, 10623 Berlin, Germany · School of Electrical Engineering and Telecommunications, University of New South Wales, Sydney, NSW, Australia
Aerial Multi-Functional RIS in Fluid Antennas-Aided Full-Duplex Networks: A Self-Optimized Hybrid Deep Reinforcement Learning Approach
Apr 15, 2026Li-Hsiang Shen, Yu-Quan ZhengHybrid Action Space RL Wireless Communications
Department of Communication Engineering, National Central University, Taoyuan 320317, Taiwan
Foundation Model-Aided Multi-Agent Reinforcement Learning for Wireless Random Access Network Optimization
Oct 6, 2026Myeung Suk Oh, Zhiyao Zhang, Alvaro Velasquez +2Wireless Resource Allocation Decentralized MARL
The Ohio State University Columbus, Ohio, USA · University of Colorado Boulder Boulder, Colorado, USA · Pathfinder Applied Research, LLC Ashburn, Virginia, USA

Sep 17, 2026J/K move · Enter open · S save

Agentic AI Networking for Heterogeneous Unmanned Aerial Systems in Low-Altitude Wireless Networks

Nguyen Duc Minh Quang, Chang Liu, Shuangyang Li, Derrick Wing Kwan Ng

School of Computing, Engineering, and Mathematical Sciences, La Trobe University, Melbourne, VIC, Australia · Chair of Communications and Information Theory, Technical University of Berlin, 10623 Berlin, Germany · School of Electrical Engineering and Telecommunications, University of New South Wales, Sydney, NSW, Australia

LLM-Guided RL Wireless Communications Wireless Resource Allocation Agentic AI Heterogeneous MARL

Open paper AskPDF arXiv

Low-altitude wireless networks (LAWNs) are emerging as a key infrastructure for heterogeneous unmanned aerial systems that support concurrent services within a shared three-dimensional airspace. Their coexistence creates strong coupling among mobility, connectivity, and shared network resources, while heterogeneous services impose distinct and time-varying requirements. These interactions naturally form a dynamic non-cooperative game in which both operating conditions and coordination objectives evolve over time. Conventional optimization and learning-based controllers typically rely on predefined objectives, limiting their ability to adapt autonomously to changing service requirements and resource priorities. To address this challenge, we propose a hierarchical hybrid large language model (LLM)- multi-agent reinforcement learning (MARL) architecture organized as a dual-loop structure. Specifically, an outer adaptation loop employs LLM-assisted game orchestration to interpret service requirements and operator intent, and reconfigure objectives and resource priorities, while an inner loop executes decentralized, parameter-conditioned MARL policies under the configured game. A logistics-monitoring case study illustrates how the proposed framework facilitates coordinated coexistence among heterogeneous services, adapting to evolving operating conditions without retraining the underlying MARL policies. Finally, we discuss key challenges and research directions toward scalable, trustworthy, and adaptive agentic LAWNs.

Abstract

Wireless Communications

Explore similar work

Agentic AI Networking for Heterogeneous Unmanned Aerial Systems in Low-Altitude Wireless Networks

Aerial Multi-Functional RIS in Fluid Antennas-Aided Full-Duplex Networks: A Self-Optimized Hybrid Deep Reinforcement Learning Approach

Foundation Model-Aided Multi-Agent Reinforcement Learning for Wireless Random Access Network Optimization