大模型与网络智能化探讨  

Discussion on Large Models and Network Intelligence

作　　者：李露 李福昌[1] Li Lu;Li Fuchang(China Unicom Research Institute,Beijing 100048,China)

机构地区：[1]中国联通研究院,北京100048

出　　处：《邮电设计技术》2025年第1期1-5,共5页Designing Techniques of Posts and Telecommunications

摘　　要：深入探讨了大模型在网络智能化中的关键作用与应用模式。大模型凭借其强大的学习能力和泛化性能,能够深度挖掘网络数据中的潜在价值,为网络运维、管理以及性能优化等多方面提供智能决策依据。同时,在架构层面提出了端边云智能协同架构,研究了相关架构特点,为网络智能化演进及支持更多大模型等新AI业务应用提供了参考。It deeply explores the key role and application modes of large models in network intelligence.Large models,with their powerful learning ability and generalization performance,can deeply mine the latent value in network data,providing intelligent decision-making support for network operation,management,and performance optimization in multiple aspects.Furthermore,it proposes an end-edge-cloud intelligent collaborative architecture at the architecture level,studies the relevant architecture features,and provides reference for the evolution of network intelligence to support new AI services.

关 键 词：大模型 人工智能 端边云协同 

分 类 号：TN929.5[电子电信—通信与信息系统]