Research on SDN Network Automated Configuration Based on Large Language Models

YANG Xing-yuan; TIAN Le; YAO Ying; PAN Fan; HU Yu-xiang

doi:10.12263/DZXB.20250463

您当前的位置：

首页 >

文章列表页 >

Research on SDN Network Automated Configuration Based on Large Language Models

Large-Scale Models and the Internet | 更新时间：2025-12-27

- Research on SDN Network Automated Configuration Based on Large Language Models
- ACTA ELECTRONICA SINICA Vol. 53, Issue 9, Pages: 3078-3088(2025)
- 作者机构：
  
  信息工程大学，河南郑州 450001
- 作者简介：
- 基金信息：
  
  National Key Research and Development Program(2022YFB2901501);Zhongyuan Science and Technology Innovation Leading Researcher Project(244200510038);Major Science and Technology Special Project of Henan Province(221100210900-02)
- DOI：10.12263/DZXB.20250463
  CLC： TP393;
- Received：31 May 2025，
  
  Accepted：11 September 2025，
  
  Published：25 September 2025
- 稿件说明：
移动端阅览
杨兴源, 田乐, 姚莹, 等. 基于大语言模型的SDN网络自动化配置研究[J]. 电子学报, 2025, 53(09): 3078-3088.

YANG Xing-yuan, TIAN Le, YAO Ying, et al. Research on SDN Network Automated Configuration Based on Large Language Models[J]. Acta Electronica Sinica, 2025, 53(09): 3078-3088.
杨兴源, 田乐, 姚莹, 等. 基于大语言模型的SDN网络自动化配置研究[J]. 电子学报, 2025, 53(09): 3078-3088. DOI：10.12263/DZXB.20250463

YANG Xing-yuan, TIAN Le, YAO Ying, et al. Research on SDN Network Automated Configuration Based on Large Language Models[J]. Acta Electronica Sinica, 2025, 53(09): 3078-3088. DOI：10.12263/DZXB.20250463

摘要

传统网络依赖人工配置，在应对规模激增、需求复杂化及实时性要求提升的现代网络环境时，效率低下且成本高昂.大语言模型（Large Language Model，LLM）凭借其出色的自然语言理解能力，在网络自动化配置中展现出巨大的潜力.面向软件定义网络（Software Defined Networking，SDN），本文提出了一种基于LLM的轻量级自动化配置方法.在数据平面，提出了一种基于检索增强生成（Retrieval-Augmented Generation，RAG）技术的代码自动生成方法RetroP4，支持基于用户意图生成P4代码；在控制平面，提出了一种基于任务分解的流表自动生成方法CtrlSynth，支持基于用户意图和数据平面P4代码生成流表配置.实验结果表明：相较于通用大模型，RetroP4生成的P4代码的语法正确性提高了25%，语义正确性提高了87.5%；CtrlSynth能够准确生成与P4代码匹配的流表信息，在流量意图不超过300条时，准确率可达100%.

Abstract

Traditional networks

which depend on manual configuration

are inefficient and expensive in the face of today’s rapidly expanding scales

increasingly complex demands

and the growing need for real-time responsiveness. Large language models (LLM)

known for their exceptional ability to understand natural language

show immense promise for automating network configurations. This paper introduces a streamlined approach to automated configuration for software defined networking (SDN)

leveraging LLM. For the data plane

we present RetroP4

a code generation method that uses retrieval-augmented generation (RAG) technology

enabling the creation of P4 code tailored to users’ intentions. In the control plane

we propose CtrlSynth

a method for automatically generating flow tables by breaking down tasks

aligning the configurations with users’ intentions and the P4 code from the data plane. Compared with general-purpose large models

the syntactic correctness of P4 code generated by RetroP4 is improved by 25%

and the semantic correctness is enhanced by 87.5%. CtrlSynth accurately produces flow table information that corresponds to the P4 code

achieving a 100% accuracy rate when dealing with up to 300 traffic-related intentions.

关键词

Keywords

references

KHORSANDROO S , SÁNCHEZ A G , TOSUN A S , et al . Hybrid SDN evolution: A comprehensive survey of the state-of-the-art [J ] . Computer Networks , 2021 , 192 : 107981 .

王朝炜 , 杜嘉楠 , 王程 , 等 . 软件定义卫星网络中基于业务调度的混合路由算法 [J ] . 电子学报 , 2024 , 52 ( 5 ): 1506 - 1515 .

WANG C W , DU J N , WANG C , et al . A hybrid routing based on traffic scheduling in double-layer software defined satellite networks [J ] . Acta Electronica Sinica , 2024 , 52 ( 5 ): 1506 - 1515 . (in Chinese)

RAMANATHAN S , ZHANG Y , GAWISH M , et al . Practical intent-driven routing configuration synthesis [C ] // Proceedings of the 20th USENIX Symposium on Networked Systems Design and Implementation . California : USENIX Association , 2023 : 629 - 644 .

HSU K F , BECKETT R , CHEN A , et al . Contra: A programmable system for performance-aware routin [C ] // Proceedings of the 17th USENIX Symposium on Networked Systems Design and Implementation . California : USENIX Association , 2020 : 701 - 721 .

BECKETT R , MAHAJAN R , MILLSTEIN T , et al . Network configuration synthesis with abstract topologies [C ] // Proceedings of the 38th ACM SIGPLAN Conference on Programming Language Design and Implementation . New York : ACM , 2017 : 437 - 451 .

TIAN B C , ZHANG X Y , ZHAI E N , et al . Safely and automatically updating in-network ACL configurations with intent language [C ] // Proceedings of the ACM Special Interest Group on Data Communication . New York : ACM , 2019 : 214 - 226 .

EL-HASSANY A , TSANKOV P , VANBEVER L , et al . NetComplete: Practical network-wide configuration synthesis with autocompletion [C ] // Proceedings of the 15th USENIX Symposium on Networked Systems Design and Implementation . California : USENIX Association , 2018 : 579 - 594 .

KOLIDES A , NAWAZ A , RATHOR A , et al . Artificial intelligence foundation and pre-trained models: Fundamentals, applications, opportunities, and social impacts [J ] . Simulation Modelling Practice and Theory , 2023 , 126 : 102754 .

FLORIDI L , CHIRIATTI M . GPT-3: Its nature, scope, limits, and consequences [J ] . Minds and Machines , 2020 , 30 ( 4 ): 681 - 694 .

LI J , TAO C Y , LI J , et al . Large language model-aware in-context learning for code generation [J ] . ACM Transactions on Software Engineering and Methodology , 2025 , 34 ( 7 ): 1 - 33 .

GAO B , HE Z M , SHARMA P , et al . Cost-efficient large language model serving for multi-turn conversations with CachedAttention [C ] // USENIX Annual Technical Conference . California : USENIX Association , 2024 : 111 - 126 .

WANG C J , SCAZZARIELLO M , FARSHIN A , et al . Making network configuration human friendly [J/OL ] . ( 2023-09-12 )[ 2025-05-24 ] . https://arxiv.org/abs/2309.06342 https://arxiv.org/abs/2309.06342 .

MONDAL R , TANG A L , BECKETT R , et al . What do LLMs need to synthesize correct router configurations? [C ] // Proceedings of the 22nd ACM Workshop on Hot Topics in Networks . New York : ACM , 2023 : 189 - 195 .

ZHANG F J , CHEN B , ZHANG Y , et al . RepoCoder: Repository-level code completion through iterative retrieval and generation [C ] // Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing . Stroudsburg : ACL , 2023 : 2471 - 2484 .

XU Z T , CRUZ M J , GUEVARA M , et al . Retrieval-augmented generation with knowledge graphs for customer service question answering [C ] // Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval . New York : ACM , 2024 : 2905 - 2909 .

WEI J , WANG X Z , SCHUURMANS D , et al . Chain-of-thought prompting elicits reasoning in large language models [C ] // Proceedings of the 36th International Conference on Neural Information Processing Systems . New York : ACM , 2022 : 24824 - 24837 .

BIRKNER R , DRACHSLER-COHEN D , VANBEVER L , et al . Config2Spec: Mining network specifications from network configurations [C ] // Proceedings of the 17th USENIX Symposium on Networked Systems Design and Implementation . California : USENIX Association , 2020 : 969 - 984 .

SUBRAMANIAN K , D'ANTONI L , AKELLA A . Genesis: Synthesizing forwarding tables in multi-tenant networks [J ] . ACM SIGPLAN Notices , 2017 , 52 ( 1 ): 572 - 585 .

DOERING N , GORLLA C , TUTTLE T , et al . Empirical analysis of efficient fine-tuning methods for large pre-trained language models [EB/OL ] . ( 2024-01-08 )[ 2025-05-21 ] . https://arXiv.org/abs/2401.04051 https://arXiv.org/abs/2401.04051 .

DING N , QIN Y J , YANG G , et al . Parameter-efficient fine-tuning of large-scale pre-trained language models [J ] . Nature Machine Intelligence , 2023 , 5 ( 3 ): 220 - 235 .

FAN S D , CONG X , FU Y P , et al . WorkflowLLM: Enhancing workflow orchestration capability of large language models [EB/OL ] . ( 2024-11-08 )[ 2025-05-21 ] . https://arXiv.org/abs/2411.05451 https://arXiv.org/abs/2411.05451 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Scalable Network Verification Technologies: State of the Art and Future

Load Balancing Based on Flow Classification for Datacenter Network

Related Author

HU Yu-xiang

PAN Fan

YAO Ying

YANG Xing-yuan

TIAN Le

HUANG Han-lin

XU Ke

LI Qi

Related Institution

University of Information Engineering

Department of Computer Science and Technology， Tsinghua University

Beijing National Research Center for Information Science and Technology

Zhongguancun Laboratory

Institute for Network Science and Cyberspace， Tsinghua University

⁰