基于知识图谱的6G网络场景认知研究  

Research on 6G network scenario cognition based on knowledge graph

作　　者：赵茁乔 承楠 陈劼[2] 陈芳炯[3] 李长乐 ZHAO Zhuoqiao;CHENG Nan;CHEN Jie;CHEN Fangjiong;LI Changle(College of Communication Engineering,Xidian University,Xi’an 710071,China;National Key Laboratory of Communication,University of Electronic Science and Technology of China,Chengdu 610054,China;School of Electronics and Information Engineering,South China University of Technology,Guangzhou 510640,China)

机构地区：[1]西安电子科技大学通信工程学院,陕西西安710071 [2]电子科技大学通信抗干扰技术国家级重点实验室,四川成都610054 [3]华南理工大学电子与信息学院,广东广州510640

出　　处：《智能科学与技术学报》2023年第4期494-504,共11页Chinese Journal of Intelligent Science and Technology

基　　金：国家重点研发计划(No.2020YFB1807700)。

摘　　要：6G网络覆盖空天地海全域,对于多样化、个性化的场景,6G网络需要提供定制化的服务,即按需服务。为实现全域全场景的按需服务,对场景特性进行精准、实时、智能的认知是一个重要前提。如何使网络自主智能地认知不同的场景及服务并将认知结果转换成场景特定的网络关键性能指标,并进一步对网络资源进行高效调度是亟待解决的关键问题。为此,将知识图谱应用于网络场景的认知识别,形成对6G网络场景的规范化描述,并构建了基于6G场景本体的知识图谱。同时,提出一种基于知识图谱嵌入的场景认知推理方法,实现了对图谱节点和关系的嵌入学习,并对场景特征节点进行了预测推理,取得了较高的准确度。该方法有助于实现6G全场景网络中场景感知、认知、按需服务的服务全生命周期自主管控,对于提高下一代网络的自主性和智能性有重要的指导意义。The 6G network covers the entire space,air,ground,and sea.For diversified and personalized scenarios,the 6G network needs to provide customized services,that is,on-demand services.In order to realize on-demand services in all domains and scenarios,accurate,real-time,and intelligent cognition of the characteristics of the scenarios is an important prerequisite.How to enable the network to autonomously and intelligently recognize different scenarios and services,convert them into scenario-specific network key performance indicator(KPI),and further efficiently schedule network resources is a key problem that urgently needs to be solved.This paper applies the knowledge graph to the cognitive recognition of network scenarios,forms a standardized description of 6G network scenarios,and builds a knowledge graph based on the 6G scenario ontology.At the same time,a scene cognition reasoning method based on knowledge graph embedding is proposed,which realizes the embedding learning of graph nodes and relationships and reasons about scene feature nodes,achieving high accuracy.The method proposed in this paper helps to realize the autonomous control of the service life cycle of scene awareness,cognition,and on-demand services in the 6G full-scenario network,and has important innovation and guiding significance for improving the autonomy and intelligence of the next-generation network.

关 键 词：6G全场景 场景认知 知识图谱 图嵌入 节点预测 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]