83 | 0 | 19 |
下载次数 | 被引频次 | 阅读次数 |
面向6G智能终端AI业务原生、融合感知、智能协同业务需求,提出了一种端边协同的6G内生智能网络架构。该架构通过分层设计(基础设施层、模型管理层、资源管控层、业务编排层),实现了端边数据管控、模型动态协同及异构资源融合调度,具备“通信+计算+数据+模型”一体化服务能力。在内生智能网络架构基础上,提出了端边智能协同系统评估模型,并围绕数据管控、模型协同、资源调度3个维度,提出了高质量数据集构建、数据管控框架、参考模型库、端边模型协同、异构资源融合管控、灵活组网等关键技术,形成端边智能协同技术体系。
Abstract:To address the requirements of artificial intelligence(AI)-native service, integrated sensing, and intelligent collaboration for 6G AI terminals, an edge-device collaborative 6G AI-native network architecture is proposed. The architecture adopts a hierarchical design(infrastructure layer, model management layer, resource control layer, and service orchestration layer) to achieve core capabilities such as edgedevice data management and control, dynamic model collaboration, and integrated heterogeneous resource scheduling, providing integrated "communication + computing + data + model" service capabilities. Based on this architecture, an edge-device AI collaborative system evaluation model is established. Focusing on three key dimensions—data management, model collaboration, and resource scheduling—key technologies are proposed, including high-quality dataset construction, data management framework, reference model library, edge-device model collaboration, heterogeneous resource integration and control, and flexible networking, forming a comprehensive edge-device intelligent collaboration technology system.
[1] IMT-2030(6G)推进组. 6G总体愿景与潜在关键技术白皮书[R]. 2021
[2] IMT-2030(6G)推进组. 6G网络架构展望白皮书[R]. 2023
[3]中国信息通信研究院.新一代智能终端蓝皮书[R]. 2024
[4] Next G Alliance. 6G technologies for wide-area cloud evolution[R]. 2023
[5] 6G-IA. White paper:European vision for the 6G network ecosystem[R]. 2024
[6] IMT-2030(6G)推进组. 6G AI即服务(AIaaS)需求研究[R]. 2023
[7] IMT-2020(5G)推进组, IMT-2030(6G)推进组.移动通信与AI融合的数据格式和模型建议书(第一阶段:物理层领域)[R]. 2023
[8] 3GPP. Study on architecture for 6G system:S2-2506096[EB/OL].(2025-05-23)[2025-06-18]. https://www. 3gpp. org/ftp/tsg_sa/WG2_Arch/TSGS2_169_Fukuoka_2025-05/Docs
[9] JIANG W, HAN H C, FENG D Q, et al. Energy-efficient and accuracy-aware DNN inference with IoT device-edge collaboration[J]. IEEE transactions on services computing, 2025,18(2):784-797. DOI:10.1109/TSC.2025.3536311
[10] LIU Z, TIAN M, DONG M, et al. MoEI:mobility-aware edge inference based on model partition and service migration[J].IEEE transactions on mobile computing, 2024, 23(10):9437-9450. DOI:10.1109/TMC.2024.3366186
[11] HE Y, FANG J, YU F R, et al. Large language models(LLMs)inference offloading and resource allocation in cloud-edge computing:an active inference approach[J]. IEEE transactions on mobile computing, 2024, 23(12):11253-11264 DOI:10.1109/TMC.2024.3415661
[12]中国移动研究院. AI+通信业务白皮书[R]. 2025
[13]中国信息通信研究院.具身智能发展报告[R]. 2024
[14]华为终端有限公司. AI终端白皮书[R]. 2025
[15] LIU Z, DU H, LIN J, et al. DNN partitioning, task offloading, and resource allocation in dynamic vehicular networks:a LyapunovGuided diffusion-based reinforcement learning approach[J].IEEE transactions on mobile computing, 2025, 24(3):1945-1962. DOI:10.1109/TMC.2024.3486728
基本信息:
DOI:
中图分类号:TP18;TN929.5
引用信息:
[1]王志勤,周吉喆,韩凯峰.端边协同的6G内生AI网络[J].中兴通讯技术,2025,31(04):29-33.
基金信息:
北京市自然科学基金资助项目(L253004); 中国科协青年人才托举工程项目(2023QNRC001,2022QNRC001)