Author Login Editor-in-Chief Peer Review Editor Work Office Work

Computer Engineering ›› 2021, Vol. 47 ›› Issue (12): 171-176,184. doi: 10.19678/j.issn.1000-3428.0059601

• Mobile Internet and Communication Technology • Previous Articles     Next Articles

Intelligent Routing Mechanism Based on Graph Neural Networks

ZHANG Peng, CHEN Bo   

  1. People's Liberation Army Strategic Support Force Information Engineering University, Zhengzhou 450002, China
  • Received:2020-09-27 Revised:2020-12-12 Published:2020-12-18

基于图神经网络的智能路由机制

张鹏, 陈博   

  1. 中国人民解放军战略支援部队信息工程大学, 郑州 450002
  • 作者简介:张鹏(1982-),男,副研究员,主研方向为网络架构、网络协议;陈博(通信作者),讲师、博士研究生。
  • 基金资助:
    国家重点研发计划“多模态智慧网络核心技术与原理平台”(2019YFB1802502)。

Abstract: Existing artificial intelligence-based routing schemes are limited in generalization performance, and fail to adapt to the topological changes of networks.This paper proposes an intelligent routing strategy named SmartRoute based on deep reinforcement learning.SmartRoute can dynamically adjust the routing strategy by sensing the network traffic distribution in real time.Additionally, it combines the topology information perception ability of graph neural network and the self-training ability of deep reinforcement learning to improve the intelligence of network routing strategy.Experimental results show that SmartRoute saves up to 9.6% of end-to-end delay, and exhibits higher robustness than DRL-TE, TIDE and other schemes.

Key words: Software-Defined Network(SDN), routing mechanism, deep reinforcement learning, Graph Neural Networks(GNN), artificial intelligence

摘要: 现有基于人工智能的路由方案泛化能力较差,难以适应动态的网络拓扑变化。提出基于深度强化学习的智能路由机制SmartRoute。通过实时感知网络中流量分布状态,动态调整路由策略,并结合图神经网络的拓扑信息感知能力和深度强化学习的自我训练能力,提升网络路由策略的智能性。实验结果表明,与DRL-TE、TIDE等方案相比,SmartRoute最多节省9.6%的端到端时延,且具有更好的鲁棒性。

关键词: 软件定义网络, 路由机制, 深度强化学习, 图神经网络, 人工智能

CLC Number: