<inline-formula><math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="block" id="M1"><mrow/></math></inline-formula>基于图学习的空天地一体化网络合作推理方法

徐煜凯; 徐波; 张凌豪; 王琴; 赵海涛

doi:null

您当前的位置：

首页 >

文章列表页 >

基于图学习的空天地一体化网络合作推理方法

更新时间：2026-03-03

- 基于图学习的空天地一体化网络合作推理方法
- Graph Learning-based Cooperative Inference Method for Space-Air-Ground Integrated Networks
- 物联网学报 2026年
- 作者机构：
  
  1.南京邮电大学通信与信息工程学院，江苏南京 210003
  2.南京邮电大学物联网学院，江苏南京 210003
- 作者简介：
  
  徐波 xubo@njupt.edu.cn
- 基金信息：
  
  国家自然科学基金(U2441226)
- DOI：
  中图分类号： TN929.5
- 收稿：2025-05-06，
  
  修回：2025-06-27，
  
  录用：2025-07-18，
- 稿件说明：
移动端阅览
徐煜凯, 徐波, 张凌豪, 等. 基于图学习的空天地一体化网络合作推理方法[J/OL]. 物联网学报, 2026.

XU Yukai, XU Bo, ZHANG Linhao, et al. Graph Learning-based Cooperative Inference Method for Space-Air-Ground Integrated Networks[J/OL]. Chinese Journal on Internet of Things, 2026.
徐煜凯, 徐波, 张凌豪, 等. 基于图学习的空天地一体化网络合作推理方法[J/OL]. 物联网学报, 2026. DOI：

XU Yukai, XU Bo, ZHANG Linhao, et al. Graph Learning-based Cooperative Inference Method for Space-Air-Ground Integrated Networks[J/OL]. Chinese Journal on Internet of Things, 2026. DOI：

基于图学习的空天地一体化网络合作推理方法

PDF下载

摘要

在空天地一体化网络 (SAGIN

space-air-ground integrated network) 中，通过推理已训练的深度神经网络 (DNN

deep neural network) 能够实现多样化的人工智能 (AI

atificial intelligence) 服务。由于DNN的参数量不断增大，在有限的通信和计算资源下，推理任务的执行过程通常需要多方的合作。本论文将认知无线电技术 (CR

cognitive radio) 融入SAGIN中，以最小化合作推理时延为目标，在卫星通信干扰阈值约束、设备服务质量约束以及发射功率约束下，以无人机 (UAV

unmanned aerial vehicle) 为推理任务的发起者建立考虑设备关联、模型分割以及功率优化的合作推理问题。该问题是多维变量耦合的非凸问题，本文根据优化变量将原问题分解，提出一种融合二次无约束二值优化 (QUBO

quadratic unconstrained binary optimization) 和注意力机制的图学习法，将目标函数和约束条件建模为QUBO形式，并且通过哈密顿量松弛生成可微的损失函数，进而利用动态多头注意力机制对图结构问题中的节点信息进行聚合，从而完成设备关联模型分割优化。此外，使用连续凸逼近 (SCA

successive convex approximation) 对发射功率进行优化，最后对三部分子问题交替迭代求解。仿真结果表明，所提方法能够将合作推理任务总时延降低11%。

Abstract

In the space-air-ground integrated network (SAGIN)

diverse atificial iintelligence (AI) applications are realized through the inference of pre-trained deep neural networks (DNNs). Owing to the ever-increasing parameter scale of DNNs

the execution of inference tasks under constrained communication and computational resources typically necessitates multi-party collaboration. This paper integrates cognitive radio (CR) technology into SAGIN

aiming to minimize inference latency while considering constraints such as satellite communication interference thresholds

device quality-of-service requirements

and transmit power limitations. We formulate a cooperative inference problem involving device association

model partitioning

and power optimization

with unmanned aerial vehicle (UAV) serving as task initiators. The resulting problem is a non-convex optimization with coupled multi-dimensional variables. Accordingly

we decompose the original problem based on optimization variables and propose a graph learning method combining quadratic unconstrained binary optimization (QUBO) and an attention mechanism. Specifically

the objective function and constraints are modeled in QUBO form

and a differentiable loss function is derived through Hamiltonian relaxation. A dynamic multi-head attention mechanism is then employed to aggregate node information in the graph-structured problem

optimizing device association and model partitioning. Additionally

transmit power is optimized using successive convex approximation (SCA). Ultimately

the three subproblems are solved alternately in an iterative process. Simulation results indicate that the proposed method reduces the total latency of collaborative inference tasks by 11%.

关键词

Keywords

references

YU S , GONG S W , SHI Q , et al . EC-SAGINs: Edge-Computing-Enhanced Space–Air–Ground-Integrated Networks for Internet of Vehicles [J ] . IEEE Internet of Things Journal , 2022 , 9 ( 8 ): 5742 - 5754 .

TAN J J , TANG F X , ZHAO M , et al . Outage Probability, Performance, and Fairness Analysis of Space–Air–Ground Integrated Network (SAGIN): UAV Altitude and Position Angle [J ] . IEEE Transactions on Wireless Communications , 2025 , 24 ( 2 ): 940 - 954 .

武艳，潘广川，姚明旿，等 . 基于深度强化学习的空天地一体化网络信息物理系统垂直切换策略 [J ] . 通信学报， 2024 , 45 ( 08 ): 180 - 191 .

WU Y , PAN G C , YAO M W , et al .. Vertical handover policy for cyber-physical systems aided by SAGIN based on deep reinforcement learning [J ] . Journal on Communications , 2024 , 45 ( 08 ): 180 - 191 .

FU S , GAO J , ZHAO L . Collaborative Multi-Resource Allocation in Terrestrial-Satellite Network Towards 6G [J ] . IEEE Transactions on Wireless Communications , 2021 , 20 ( 11 ): 7057 - 7071 .

LEE J-H , PARK J , BENNIS M , et al . Integrating LEO Satellites and Multi-UAV Reinforcement Learning for Hybrid FSO/RF Non-Terrestrial Networks [J ] . IEEE Transactions on Vehicular Technology , 2023 , 72 ( 3 ): 3647 - 3662 .

CHEN R R , SUN Y J , LIANG L P , et al . Joint Power Allocation and Placement Scheme for UAV-Assisted IoT With QoS Guarantee [J ] . IEEE Transactions on Vehicular Technology , 2022 , 71 ( 1 ): 1066 - 1071 .

QU K G , ZHUANG W H , WU W , et al . Stochastic Cumulative DNN Inference With RL-Aided Adaptive IoT Device-Edge Collaboration [J ] . IEEE Internet of Things Journal , 2023 , 10 ( 20 ): 18000 - 18015 .

XU X D , YAN K W , HAN S J , et al . Learning-Based Edge-Device Collaborative DNN Inference in IoVT Networks [J ] . IEEE Internet of Things Journal , 2024 , 11 ( 5 ): 7989 - 8004 .

OTTER D W , MEDINA J R , KALITA J K . A Survey of the Usages of Deep Learning for Natural Language Processing [J ] . IEEE Transactions on Neural Networks and Learning Systems , 2021 , 32 ( 2 ): 604 - 624 .

Bakambekova A , Kouzayha N , Al-Naffouri T . On the Interplay of Artificial Intelligence and Space-Air-Ground Integrated Networks: A Survey [J ] . IEEE Open Journal of the Communications Society , 2024 , 5 : 4613 - 4673 .

王俊华，罗菲，高广鑫，等 . 动态车辆网络场景中的协同空地计算卸载和资源优化 [J ] . 电子与信息学报， 2025 , 47 ( 01 ): 102 - 115 .

WANG J H , Luo F , GAO G X , et al . Collaborative Air-Ground Computation Offloading and ResourceOptimization in Dynamic Vehicular Network Scenarios [J ] . Journal of Electronics & Information Technology , 2025 , 47 ( 01 ): 102 - 115 .

HE W C , GUO S Y , GUO S , et al . Joint DNN Partition Deployment and Resource Allocation for Delay-Sensitive Deep Learning Inference in IoT [J ] . IEEE Internet of Things Journal , 2020 , 7 ( 10 ): 9241 - 9254 .

YAN J , BI S Z , Y. ZHANG Y-J A . Optimal Model Placement and Online Model Splitting for Device-Edge Co-Inference [J ] . IEEE Transactions on Wireless Communications , 2022 , 21 ( 10 ): 8354 - 8367 .

HAN S J , ZHANG W Z , XU X D , et al . S2E-DECI: Secrecy and Energy-Efficient Dual-Aware Device-Edge Co-Inference for AIoT [J ] . IEEE Internet of Things Journal , 2024 , 11 ( 24 ): 39142 - 39157 .

LIU C X , FENG W , CHEN Y F , el at . Cell-free satellite-UAV networks for 6G wide-area Internet of Things [J ] . IEEE Journal on Selected Areas in Communications , 2021 , 39 ( 4 ): 1116 - 1131 .

WANG Z C , ZHANG L , FENG D Q , et al . Intelligent Cloud-Edge Collaborations for Energy-Efficient User Association and Power Allocation in Space-Air-Ground Integrated Networks [J ] . IEEE Journal on Selected Areas in Communications , 2024 , 42 ( 12 ): 3659 - 3673 .

WANG Y F , ZHAO L Q , CHU X L , et al . E2E Network Slicing Optimization for Control- and User-Plane Separation-Based SAGIN With DRL [J ] . IEEE Transactions on Vehicular Technology , 2024 , 73 ( 11 ): 16680 - 16696 .

ZHOU Z Z , ZHANG Q Q , GE J G , et al . Hierarchical Cognitive Spectrum Sharing in Space-Air-Ground Integrated Networks [J ] . IEEE Transactions on Wireless Communications , 2024 , 24 ( 2 ): 1430 - 1447 .

NASERI P , MATOS S A , COSTA J R , et al . Dual-Band Dual-Linear-to-Circular Polarization Converter in Transmission Mode Application to K/Ka -Band Satellite Communications [J ] . IEEE Transactions on Antennas and Propagation , 2018 , 66 ( 12 ): 7128 - 7137 .

逄小玮，蒋旭，卢华兵，等 . 面向6G多维扩展的新型多址接入技术综述 [J ] . 电子与信息学报， 2024 , 46 ( 06 ): 2323 - 2334 .

PANG X W , JIANG X , LU H B , et al . An Overview of Novel Multi-access Techniques for Multi-dimensional Expanded 6G [J ] . Journal of Electronics & Information Technology , 2024 , 46 ( 06 ): 2323 - 2334 .

LIU R , GUO K F , AN K , et al . Resource Allocation for Cognitive Satellite-HAP-Terrestrial Networks With Non-Orthogonal Multiple Access [J ] . IEEE Transactions on Vehicular Technology , 2023 , 72 ( 7 ): 9659 - 9663 .

GUO S A , ZHAO X H . Multi-Agent Deep Reinforcement Learning Based Transmission Latency Minimization for Delay-Sensitive Cognitive Satellite-UAV Networks [J ] . IEEE Transactions on Communications , 2023 , 71 ( 1 ): 131 - 144 .

PERVEZ F , ZHAO L , YANG C G . Joint User Association, Power Optimization and Trajectory Control in an Integrated Satellite-Aerial-Terrestrial Network [J ] . IEEE Transactions on Wireless Communications , 2022 , 21 ( 5 ): 3279 - 3290 .

SHI W Q , ZHOU S , NIU Z S , et al . Multiuser Co-Inference With Batch Processing Capable Edge Server [J ] . IEEE Transactions on Wireless Communications , 2023 , 22 ( 1 ): 286 - 300 .

WU Q Q , ZENG Y , ZHANG R . Joint Trajectory and Communication Design for Multi-UAV Enabled Wireless Networks [J ] . IEEE Transactions on Wireless Communications , 2018 , 17 ( 3 ): 2109 - 2121 .

ABBASI O , YANIKOMEROGLU H , EBRAHIMI A , et al . Trajectory Design and Power Allocation for Drone-Assisted NR-V2X Network With Dynamic NOMA/OMA [J ] . IEEE Transactions on Wireless Communications , 2020 , 19 ( 11 ): 7153 - 7168 .

GUO K F , AN K , ZHANG B N , et al . Physical Layer Security for Multiuser Satellite Communication Systems With Threshold-Based Scheduling Scheme [J ] . IEEE Transactions on Vehicular Technology , 2020 , 69 ( 5 ): 5129 - 5141 .

SCHUETZ M J A , BRUBAKER J K , KATZGRABER H G . Combinatorial optimization with physics-inspired graph neural networks [J ] . Nature Machine Intelligence , 2022 , 4 ( 4 ): 367 - 377 .

ATOBE Y , TAWADA M , TOGAWA N . Hybrid Annealing Method Based on subQUBO Model Extraction With Multiple Solution Instances [J ] . IEEE Transactions on Computers , 2022 , 71 ( 10 ): 2606 - 2619 .

HU B , GUO K H , WANG X K , et al . RRL-GAT: Graph Attention Network-Driven Multilabel Image Robust Representation Learning [J ] . IEEE Internet of Things Journal , 2022 , 9 ( 12 ): 9167 - 9178 .

LI Z W , WANG Q K , YUAN C A , et al . Predicting MiRNA-Disease Associations by Graph Representation Learning Based on Jumping Knowledge Networks [J ] . IEEE/ACM Transactions on Computational Biology and Bioinformatics , 2023 , 20 ( 5 ): 2629 - 2638 .

He J C , Cheng N , Yin Z S , et al . Load-Aware Network Resource Orchestration in LEO Satellite Network: A GAT-Based Approach [J ] . IEEE Internet of Things Journal , 2024 , 11 ( 9 ): 15969 - 15984 .

Zeng Y P , Chen S S , Li J D , et al . Online Optimization in UAV-Enabled MEC System: Minimizing Long-Term Energy Consumption Under Adapting to Heterogeneous Demands [J ] . IEEE Internet of Things Journal , 2024 , 11 ( 9 ): 32143 - 32159 .

Han D-J , Fang W Z , Hosseinalipour S , et al . Orchestrating Federated Learning in Space-Air- Ground Integrated Networks: Adaptive Data Offloading and Seamless Handover [J ] . IEEE Journal on Selected Areas in Communications , 2024 , 42 ( 12 ): 3505 - 3520 .

ZHANG X Y , ZOU J H , HE K M , et al . Accelerating Very Deep Convolutional Networks for Classification and Detection [J ] . IEEE Transactions on Pattern Analysis and Machine Intelligence , 2016 , 38 ( 10 ): 1943 - 1955 .

WANG Y T , QIAN Z , HE L J , et al . Intelligent Online Computation Offloading for Wireless-Powered Mobile-Edge Computing [J ] . IEEE Internet of Things Journal , 2024 , 11 ( 17 ): 28960 - 28974 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

暂无数据

基于图学习的空天地一体化网络合作推理方法

Graph Learning-based Cooperative Inference Method for Space-Air-Ground Integrated Networks

DOI：

摘要

Abstract

关键词

Keywords

references