6G智慧底座:AI内生光网络架构与关键技术  

6G Smart Infrastructure:AI-Native Optical Network Architecture and Key Technologies

作　　者：李欣悦 顾仁涛[1] 纪越峰[1] LI Xinyue;GU Rentao;JI Yuefeng(Beijing University of Posts and Telecommunications,Beijing 100876,China)

机构地区：[1]北京邮电大学,北京100876

出　　处：《移动通信》2024年第7期15-20,27,共7页Mobile Communications

基　　金：国家自然科学基金企业创新发展联合基金重点支持类项目“大规模宽谱域光网络的精简建模与智能调控”(U21B2005)。

摘　　要：“内生智慧”是6G网络架构的关键特征,需要将数据、算力、算法、智能融合形成以AI为内生能力的新一代通信系统。大规模宽谱域光网络作为承载业务的核心设施,AI内生的光网络将为6G网络提供基础和支持。首先,阐述了光网络智能化的发展历程,明确了AI内生光网络的关键特征,分析了其在传输、组网、业务三个方面的挑战。其次,提出了一种AI内生的光网络架构,通过“四层双闭环”的协同实现光网络的AI内生。之后,讨论了AI内生光网络的关键技术,包括了数字孪生技术、算网融合技术与智能泛在技术,促进光网络智能内生化、管控一体化、服务AI化,为6G实现“智能普惠”构建智慧底座。“Endogenous intelligence”is a key feature of 6G network architecture,necessitating the integration of data,computing power,algorithms,and intelligence to form a new generation of AI-native communication systems.Large-scale broadband optical networks,as the core infrastructure for service delivery,will provide the foundation and support for 6G networks.This paper first elucidates the development of intelligent optical networks,identifying the key characteristics of AI-native optical networks and analyzing their challenges in transmission,networking,and service provision.It then proposes an AI-native optical network architecture,achieved through a“four-layer dual-closed-loop”collaboration.Furthermore,the paper discusses the critical technologies for AI-native optical networks,including digital twin technology,computing-network convergence,and pervasive intelligence.These technologies promote the endogenous intelligence,integrated management,and AI-driven services of optical networks,establishing the smart infrastructure for realizing“intelligent universal accessibility”in 6G networks.

关 键 词：6G 光网络 内生智能 人工智能 

分 类 号：TN929.1[电子电信—通信与信息系统]