GAI使能网络智能化：基于LLM的网络操作系统

霍如; 沙宗轩; 吕科呈; 陈伟; 汪硕; 黄韬; F. Richard Yu

doi:10.11959/j.issn.1000-436x.2025113

您当前的位置：

首页 >

文章列表页 >

GAI使能网络智能化：基于LLM的网络操作系统

学术论文 | 更新时间：2025-07-04

- GAI使能网络智能化：基于LLM的网络操作系统
- General artificial intelligence enables network intelligence: a network operating system based on large language model
- 通信学报 2025年46卷第6期页码：32-44
- 作者机构：
  
  1.北京工业大学信息科学技术学院，北京 100124
  2.紫金山实验室，江苏南京 211111
  3.北京邮电大学网络与交换国家重点实验室，北京 100876
  4.卡尔顿大学，渥太华 K1S 5B6
- 作者简介：
  
  [ "霍如（1988- ），女，黑龙江哈尔滨人，博士，北京工业大学副教授，主要研究方向为未来网络、网络智能化、区块链、资源调度等。" ]
  沙宗轩（1990- ），男，回族，安徽蚌埠人，北京工业大学博士生，主要研究方向为未来网络、网络人工智能、深度学习、强化学习等。
  吕科呈（2001- ），男，广西玉林人，北京工业大学硕士生，主要研究方向为未来网络、车联网、资源调度等。
  陈伟（1996- ），男，河北三河人，北京邮电大学博士生，主要研究方向为边缘计算、数字孪生、区块链等。
  汪硕（1991- ），男，河南灵宝人，博士，北京邮电大学副教授，主要研究方向为数据中心网络、软件定义网络、网络流量调度等。
  黄韬（1980- ），男，重庆人，博士，北京邮电大学教授，主要研究方向为未来网络体系架构、软件定义网络、网络虚拟化等。
  F. Richard Yu（1974- ），男，加拿大工程院院士，卡尔顿大学教授，主要研究方向为互联网自主智能、自动驾驶、网络空间安全等。
- 基金信息：
  
  北京市自然科学基金资助项目(4254064)
- DOI：10.11959/j.issn.1000-436x.2025113
  中图分类号： TP393
- 收稿：2025-04-28，
  
  修回：2025-06-09，
  
  纸质出版：2025-06-25
- 稿件说明：
移动端阅览
霍如,沙宗轩,吕科呈等.,GAI使能网络智能化：基于,LLM的网络操作系统[J].通信学报,2025,46(06):32-44.

HUO Ru,SHA Zongxuan,LYU Kecheng,et al.General artificial intelligence enables network intelligence: a network operating system based on large language model[J].Journal on Communications,2025,46(06):32-44.
霍如,沙宗轩,吕科呈等.,GAI使能网络智能化：基于,LLM的网络操作系统[J].通信学报,2025,46(06):32-44. DOI： 10.11959/j.issn.1000-436x.2025113.

HUO Ru,SHA Zongxuan,LYU Kecheng,et al.General artificial intelligence enables network intelligence: a network operating system based on large language model[J].Journal on Communications,2025,46(06):32-44. DOI： 10.11959/j.issn.1000-436x.2025113.

摘要

为了实现对网络高智能化可管可控的目标，提出了基于大语言模型（LLM）的高可控大网级网络操作系统。通过网络态势感知和微调大模型的意图理解优化资源配置，实现智能化运维。同时，设计了面向中国网络操作系统（CNOS）的网络大模型微调和推理流程，可识别CNOS指令并对系统反馈结果进行归纳总结，周期性的自主训练及模型迭代更新。实验结果表明，所提系统能够快速识别并准确转换用户指令，有效降低操作系统管理任务时间，实现网络资源的智能化调度和配置，提升网络操作系统的可控性和人机交互的友好程度。

Abstract

To achieve the goal of highly intelligent and controllable network

and a highly controllable large-scale network operating system was proposed based on large language model (LLM). Through network situation awareness and the intention understanding of fine-tuning LLM

resource allocation was optimized to achieve intelligent operation and maintenance. Meanwhile

the fine-tuning and inference flow for China network operation system (CNOS) oriented network large model was designed to recognize the CNOS system instructions and summarize the feedback results of the system

as well as periodic autonomous training and iterative model updating. The experimental results show that the proposed system can quickly identify and accurately convert user instructions

effectively reduce the management task time of the operating system

realize the intelligent scheduling and configuration of network resources

and improve the controllability of the network operating system and the friendliness of human-computer interaction.

关键词

Keywords

references

WANG Y T , PAN Y H , YAN M , et al . A survey on ChatGPT: AI-generated contents, challenges, and solutions [J ] . IEEE Open Journal of the Computer Society , 2023 , 4 : 280 - 302 .

WU J Y , GAN W S , CHEN Z F , et al . AI-generated content (AIGC): a survey [J ] . arXiv Preprint , arXiv: 2304.06632 , 2023 .

WEI J , WANG X Z , SCHUURMANS D , et al . Chain-of-thought prompting elicits reasoning in large language models [J ] . Advances in Neural Information Processing Systems , 2022 , 35 : 24824 - 24837 .

LONG O , WU J , XU J , et al . Training language models to follow instructions with human feedback [J ] . Advances in Neural Information Processing Systems , 2022 , 35 : 27730 - 27744 .

黄韬 , 刘江 , 汪硕 , 等 . 未来网络技术与发展趋势综述 [J ] . 通信学报 , 2021 , 42 ( 1 ): 130 - 150 .

HUANG T , LIU J , WANG S , et al . Survey of the future network technology and trend [J ] . Journal on Communications , 2021 , 42 ( 1 ): 130 - 150 .

黄韬 , 谈沙 , 谢人超 , 等 . 网络操作系统的研究进展与展望 [J ] . 北京邮电大学学报 , 2024 , 47 ( 2 ): 1 - 10 .

HUANG T , TAN S , XIE R C , et al . A review of and prospects for network operating system research [J ] . Journal of Beijing University of Posts and Telecommunications , 2024 , 47 ( 2 ): 1 - 10 .

KHALIDI Y . SONiC: the networking switch software that powers the microsoft global cloud [R ] . 2017 .

CHOI S , BURKOV B , ECKERT A , et al . FBOSS: building switch software at scale [C ] // Proceedings of the 2018 Conference of the ACM Special Interest Group on Data Communication . New York : ACM Press , 2018 : 342 - 356 .

RAIAAN M A K , MUKTA M S H , FATEMA K , et al . A review on large language models: architectures, applications, taxonomies, open issues and challenges [J ] . IEEE Access , 2024 , 12 : 26839 - 26874 .

WANG J J , HUANG Y C , CHEN C Y , et al . Software testing with large language models: survey, landscape, and vision [J ] . IEEE Transactions on Software Engineering , 2024 , 50 ( 4 ): 911 - 936 .

BROWN T , MANN B , RYDER N , et al . Language models are few-shot learners [J ] . Advances in Neural Information Processing Systems , 2020 , 33 : 1877 - 1901 .

ZHANG S S , ROLLER S , GOYAL N , et al . OPT: open pre-trained transformer language models [J ] . arxiv Preprint , arxiv: 2205.01068 , 2022 .

AAKANKSHA C , SHARAN N , JACOB D , et al . PaLM: scaling language modeling with pathways [J ] . Journal of Machine Learning Research , 2023 , 24 : 1 - 113 .

CORONADO E , BEHRAVESH R , SUBRAMANYA T , et al . Zero touch management: a survey of network automation solutions for 5G and 6G networks [J ] . IEEE Communications Surveys & Tutorials , 2022 , 24 ( 4 ): 2535 - 2578 .

BARIAH L , ZOU H , ZHAO Q Y , et al . Understanding telecom language through large language models [C ] // Proceedings of the GLOBECOM 2023 - 2023 IEEE Global Communications Conference . Piscataway : IEEE Press , 2023 : 6542 - 6547 .

KHOLGH D K , KOSTAKOS P . PAC-GPT: a novel approach to generating synthetic network traffic with GPT-3 [J ] . IEEE Access , 2023 , 11 : 114936 - 114951 .

GUPTA M , AKIRI C , ARYAL K , et al . From ChatGPT to ThreatGPT: impact of generative AI in cybersecurity and privacy [J ] . IEEE Access , 2023 , 11 : 80218 - 80245 .

SHARMA P , YEGNESWARAN V . PROSPER: extracting protocol specifications using large language models [C ] // Proceedings of the 22nd ACM Workshop on Hot Topics in Networks . New York : ACM Press , 2023 : 41 - 47 .

CHATZISTEFANIDIS I , LEONE A , NIKAEIN N . Maestro: LLM-driven collaborative automation of intent-based 6G networks [J ] . IEEE Networking Letters , 2024 , 6 ( 4 ): 227 - 231 .

LIRA O G , CAICEDO O M , DA FONSECA N L S . Large language models for zero touch network configuration management [J ] . IEEE Communications Magazine , 2024 : 1 - 8 .

MEKRACHE A , KSENTINI A , VERIKOUKIS C . Intent-based management of next-generation networks: an LLM-centric approach [J ] . IEEE Network , 2024 , 38 ( 5 ): 29 - 36 .

HUANG Y D , XU M R , ZHANG X Y , et al . AI-generated network design: a diffusion model-based learning approach [J ] . IEEE Network , 2024 , 38 ( 3 ): 202 - 209 .

浏览量

1376

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

网络行为孪生驱动的物联网异常流量检测

多芯粒大模型加速器推理协同优化方法

大语言模型赋能的网络配置分析技术研究

ResilM-IBN：一种自愈型多智能体网络管理框架

基于元特征映射与经验黑翅鸢优化的自动化网络流量预测框架