Skip to content
Preprint

Dual-Layer Over-the-Air Federated Learning in LEO Satellite Networks: Architecture, Key Technologies and Applications

Aug 2026 · 0 citations · 15 references
Engineering

TL;DR

This article outlines the fundamental principles of the dual-layer OTA model and introduces the adaptive BH mechanism designed for time-varying topologies, aiming to provide insights for the evolution of ubiquitous non-terrestrial intelligence.

Abstract

Low Earth orbit (LEO) satellite networks are emerging as a pivotal infrastructure for global edge intelligence. In this context, integrating over-the-air (OTA) computation with adaptive beam hopping (BH) provides an innovative framework that seamlessly merges physical-layer analog aggregation with dynamic resource orchestration. This effectively overcomes the stringent bandwidth and power constraints of space platforms while extending federated learning (FL) to pervasive Internet-of-things (IoT) deployments. In this article, we first outline the fundamental principles of the dual-layer OTA model and introduce the adaptive BH mechanism designed for time-varying topologies. Then, we summarize the distinct advantages of this learning-centric architecture, which include decoupling aggregation latency from device density, optimizing spatio-temporal resource efficiency, and balancing data freshness with channel quality. Several application scenarios are explored to highlight the framework's potential across diverse vertical industries. Furthermore, a specific case is studied to demonstrate the practical efficacy of the proposed scheduling policy. The results reveal substantial performance gains in terms of model convergence speed and data utilization for satellite-based FL systems. Finally, we discuss the implementation challenges and outline future research directions, aiming to provide insights for the evolution of ubiquitous non-terrestrial intelligence.

View source

Similar papers

Conference Jul 2026

Adaptive Open RAN Data Offloading in Low Earth Orbit Satellite Networks

The Open Radio Access Network (O-RAN) paradigm, with its open interfaces and intelligent functions, is a key enabler for next-generation wireless systems. We investigate the deployment of O-RAN-based network slice functions over Low Earth Orbit (LEO) satellite networks with Mobile Edge Computing (MEC) capabilities. To provide energy-efficient and low-latency services through distributed data processing, we formulate a slice function data offloading problem aimed at jointly optimizing end-to-end (E2E) latency and energy consumption. We model the problem as an MDP and propose a Deep Reinforcement Learning (DRL)-based solution. The proposed DRL agent learns efficient offloading policies by balancing computation and communication costs in the dynamic satellite environment. Simulation results show that our DRL-based approach significantly outperforms conventional benchmarks, achieving enhanced latency and energy performance, enabling intelligent orchestration of O-RAN slices over LEO satellite networks.

S. Shinde, Daniele Tarchi, Carlo Fischione · 0 citations
Conference Jul 2026

LLM-Assisted Network Management for Digital Twin-Enabled 5G Systems

Efficient long-term network evolution is becoming increasingly critical in dense 5G-Advanced and beyond cellular systems, where persistent traffic imbalances and localized congestion pose significant challenges that conventional short-term radio resource management alone cannot fully mitigate. This paper proposes a digital twin (DT)-enabled non-real-time (NRT) network evolution framework integrated with a large language model (LLM). Within this architecture, the digital twin provides a high-fidelity, controllable environment for evaluating infrastructure actions, while the LLM serves as a strategic orchestration engine that recommends cost-efficient network upgrades based on observed network states. Unlike traditional optimization methods that require exhaustive mathematical reformulations for each specific scenario, the proposed framework leverages the reasoning capabilities of LLMs to interpret operator objectives and constraints in natural language, generating structured evolution plans. The considered NRT action space encompasses antenna upgrades, bandwidth expansion, and new base station (BS) deployment. A techno-economic formulation is introduced to jointly evaluate load reduction performance and overall economic expenditure. Numerical results in a dense cellular scenario demonstrate that the framework effectively reduces peak resource utilization and provides diverse, coordinated evolution strategies tailored to varying network conditions.

Yukai Wang, Janghee Woo, G. Hahm et al. · 0 citations
Conference Jul 2026

Unified Fog Computing Architecture Integrating O-RAN, MEC, and OneM2M

Fifth-generation (5G) networks and the Internet of Things (IoT) demand unprecedented levels of scalability and ultra-low latency. Addressing these needs requires not only advanced radio technologies but also a cohesive integration of diverse architectural standards. In this paper, we present a unified fog computing framework inspired by ETSI and OneM2M technical literature that merges the Open Radio Access Network (O-RAN) architecture, Multi-Access Edge Computing (MEC), and the OneM2M IoT standard. This approach enables real-time resource allocation, reduces end-to-end latency, alleviates network congestion, and streamlines interoperability across heterogeneous deployments. Through a simulated testbed, we demonstrate how MEC and OneM2M elements can use near-RT RIC intel to optimize service delivery. The results highlight the feasibility and potential performance gains of an integrated O-RAN-MEC-OneM2M environment, paving the way for more robust, scalable, and efficient 5G IoT solutions.

Ramon A. S. Carvalho, Fuad M. Abinader, Thiago S. da Silva · 0 citations
Open access 2026

NTN-Aware Federated Learning Framework With Hierarchical Satellite Aggregation and LoRA-Based Parameter Efficiency

Nonterrestrial networks (NTNs) based on low-Earth-orbit (LEO) satellite constellations provide promising platforms for conducting global-scale federated learning (FL). However, a fundamental feasibility barrier remains: transmitting full-model updates often exceeds the typical LEO visibility windows (30–90 s), resulting in systematic client dropouts and unstable training processes. The existing approaches largely treat communication efficiency and satellite topologies independently, leaving this challenge unresolved. In this paper, we propose a unified NTN-aware FL framework that integrates low-rank adaptation (LoRA) with a three-tier hierarchical aggregation architecture. We show that LoRA resolves the feasibility barrier not only through incremental compression but also by reducing the uplink payload size by <inline-formula> <tex-math notation="LaTeX">$74.6\times $ </tex-math></inline-formula>, thereby shifting the system bottleneck from communication-limited operations to computation-limited operations. This shift enables a hierarchical aggregation scheme that is otherwise infeasible under LEO visibility constraints. The proposed three-tier protocol combines satellite-side weighted aggregation, intersatellite link (ISL)-averaging consensus, and gateway-level global aggregation. To address intermittent connectivity issues, we further develop a staleness-aware asynchronous extension with a satellite-tailored discount function. In addition, personalized LoRA adapters enable client-specific adaptations to be implemented under heterogeneous channel conditions. We establish a rigorous NTN system model that captures topology dynamics and visibility constraints and prove its convergence under standard nonconvex assumptions. Simulations performed under realistic NTN settings demonstrate that the proposed method achieves <inline-formula> <tex-math notation="LaTeX">${R} ^{2}$ </tex-math></inline-formula> = 91.3%, closely matching full-model FL (91.8%) with a <inline-formula> <tex-math notation="LaTeX">$74.6\times $ </tex-math></inline-formula> parameter reduction. In terms of latency, the uplink contribution is reduced from 55.8% to 4.0%, while client-side computations account for 93.8% of the end-to-end latency. In terms of reliability, the client dropout rate is lowered from 65.3% to 25.4%. Regarding efficiency, ISL traffic is reduced by <inline-formula> <tex-math notation="LaTeX">$23\times $ </tex-math></inline-formula> while achieving sublinear round completion scaling up to 10,000 clients.

Muhammad Shoaib Ayub, A. Khan, Felipe Augusto Pereira et al. · 0 citations
2026

Lifelong Learning-Based SDN Design for Dynamic Configuration and Resource Allocation in Satellite–Terrestrial Networks

Satellite-airborne-terrestrial edge computing networks (SATECNs) emerge as a global solution for Internet of Things (IoT) since they can provide global coverage even in remote areas and under natural disasters. However, their dynamic and non-stationary nature makes control and resource allocation more challenging. Preserving data freshness is crucial in many IoT applications and requires timely decisions. To address these challenges, we present a knowledge-base software-defined networking architecture for satellite–airborne–terrestrial networks (KB-SAT-SDN) that enables collaboration between SDN controllers to optimize SAT configurations. A shared knowledge base (KB) is built through lifelong learning (LL) to continuously adapt and efficiently manage computing and networking resources to minimize the age of information (AoI) and energy consumption. To further accelerate learning, we exploit the heterogeneity of nodes and offloading decisions by defining different learning domains and designing a cross-domain lifelong learning (CDLL-SATECN) algorithm. With domain-specific projections, knowledge is shared between domains. Numerical results show that CDLL reduces average AoI and energy by up to 70% and converges $8\times $ faster than existing baselines. It achieves the lowest or near-lowest penalty across all RL domains, nearly halving Natural Actor-Critic (NAC)’s penalty in the most complex domains. LEO assistance lowers penalty/AoI from 61.9/49.4 to 48.5/45.1 relative to a domain without LEO, while reducing UAV energy and queues. The sensitivity analysis confirms that CDLL maintains a stable AoI–energy tradeoff over a broad range of weighting parameters.

Yinxuan Wu, Ning Wang, B. Lorenzo et al. · 0 citations
Review Jul 2026

Sustainable Air-Ground Integrated Coverage Networks: ISCC Architecture, Technologies, and Testbed

The rapid emergence of sixth-generation (6G) networks and the low-altitude economy has accelerated the evolution of wireless infrastructures toward air-ground integrated coverage networks (AGICNs), which seamlessly fuse terrestrial and aerial communication resources. However, existing AGICN studies primarily focus on coverage enhancement, while ignoring sustainability. Pursuing sustainable AGICNs introduces new challenges due to the multidimensional resource coupling across heterogeneous air-ground segments. In view of this, this paper presents a comprehensive survey and tutorial on sustainable AGICNs, aiming to balance coverage capacity with carbon efficiency in low-altitude economies. An integrated sensing, communication, and computation (ISCC)-driven architecture, which enables dynamic resource orchestration through closed-loop control, is proposed. We thus introduce a multi-dimensional sustainability metric system, which covers operational efficiency, task-oriented performance, and full lifecycle carbon emissions, to quantify energy and carbon footprints. We review enabling technologies, including artificial intelligence, hybrid precoding, integrated sensing and communication, and simultaneous wireless information and power transfer, and discuss their integration into the ISCC framework to minimize energy consumption while maintaining robust coverage. Experimental results on a real-world testbed demonstrate a 20% reduction in power consumption while achieving over 90% coverage probability, highlighting the feasibility of sustainable AGICNs for future green networks.

J. Liu, X. Zhang, M. Sheng et al. · 0 citations