6G无线网络内生AI设计及端网协同技术研究OA
Research on 6G Wireless Network Native AI Design and Network-Terminal Collaboration Technologies
为适配轻量化智能体终端密集接入场景,解决传统外挂式AI架构决策冗余、端网协同不足的问题,结合6G场景与需求,以及3GPP 6G标准化节奏,开展无线网络内生AI架构及端网协同技术研究.通过梳理典型场景的技术需求,提炼内生AI的核心设计原则,构建无线网络内生AI的逻辑架构,以逻辑架构为基础提出端网协同的技术体系,并分析了对3GPP 6G标准设计的影响,可为后续3GPP 6G标准制定提供技术参考.
To adapt to dense access scenario of lightweight AI-agent terminals and solve the problems of redundant decision-making and insufficient network-terminal collaboration in traditional external AI architecture,the research is performed on wireless network native AI architecture and network-terminal collaboration technologies,with the incorporation of 6G scenarios and requirements as well as 3GPP 6G standardization roadmap.By sorting out the technical requirements of typical scenarios,the core design principles of native AI are refined,a wireless network native AI logical architecture is constructed,and then a network-terminal collaboration technology system is designed base on this architecture,and the impact on 3GPP 6G standardization design is analyzed.This work will provide technical reference for the specification of 3GPP 6G standards.
刘建军;佟舟;王爱玲;李娜;刘光毅
中国移动通信有限公司研究院,北京 100032||中关村泛联移动通信技术创新应用研究院,北京 100088中国移动通信有限公司研究院,北京 100032||中关村泛联移动通信技术创新应用研究院,北京 100088中国移动通信有限公司研究院,北京 100032中国移动通信有限公司研究院,北京 100032||中关村泛联移动通信技术创新应用研究院,北京 100088中国移动通信有限公司研究院,北京 100032||中关村泛联移动通信技术创新应用研究院,北京 100088
信息技术与安全科学
6G无线网络内生智能端网协同轻量化终端动态算力卸载智能体
6G wireless networknative artificial intelligencenetwork-terminal collaborationlightweight terminaldynamic computing offloadingAI agent
《移动通信》 2026 (8)
9-16,8
国家重点研发计划重点专项(2024YFE0200600)
评论