新型电力系统中数字孪生辅助的网络切片资源分配
CSTR:
作者:
作者单位:

(华北电力大学新能源电力系统国家重点实验室,北京 102206)

作者简介:

唐奕明(2001-),男,硕士研究生,研究方向为电力系统暂态稳定性分析。

通讯作者:

中图分类号:

TN915;TM734

基金项目:


Software-Defined Network Slicing Test Bench and Validation for White-Box Switches
Author:
Affiliation:

(State Key Laboratory of Alternate Electrical Power System with Renewable Energy Source,North China Electric Power University,Beijing 102206,CHN)

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对新型电力系统中服务功能链(SFC)部署面临的资源需求动态变化与低时延需求,文章提出一种数字孪生辅助的虚拟网络功能(VNF)映射与关联优化算法。首先,通过构建 VNF数字孪生体实时采集资源需求信息,建立以最小化 SFC端到端时延和 VNF平均数字孪生同步时延为目标的马尔科夫决策过程(MDP)模型,并考虑计算、存储及带宽资源约束。其次,将原问题分解为 VNF映射与孪生服务节点关联两个阶段,分别提出基于分布式近端策略优化(DPPO)的 VNF映射算法和基于优势演员 -评论家(A2C)的孪生服务节点关联算法,以应对大规模 SFC场景下的动作空间爆炸与策略波动问题。仿真结果表明,所提算法能够有效降低 SFC端到端时延及 VNF数字孪生同步时延,在多业务动态环境下具备良好的收敛性与资源适配能力。

    Abstract:

    The deployment of service function chains(SFCs)in modern power systems faces considerable challenges owing to dynamically changing resource demands and stringent low-latency requirements. To address these challenges, this paper presents a digital twin-assisted optimization algorithm for virtual network function(VNF)mapping and association. Initially, VNF digital twins were constructed to capture real-time resource demand information, thus enabling the formulation of a Markov decision process(MDP)model. The optimization objective was to minimize both the end-to-end delay of the SFCs and the average synchronization delay of the VNF digital twins while satisfying the constraints on computing,storage,and bandwidth resources. Given the sequential nature of VNF mapping and twin service node association,the original problem was decomposed into two stages. A distributed proximal policy optimization (DPPO)-based algorithm was proposed for the VNF mapping stage,and an advantage actor-critic(A2C)-based algorithm was developed for the twin-service node association stage. These algorithms effectively mitigate the issues of action-space explosion and considerable policy fluctuations encountered by heuristic and basic reinforcement learning methods as the number of SFCs increases. The simulation results demonstrate that the proposed DPPO-A2C algorithm can substantially reduce the SFC end-to-end and VNF digital twin synchronization delays by designing appropriate mapping and association schemes, exhibiting superior convergence and resource adaptation performance in dynamic multiservice scenarios.

    参考文献
    相似文献
    引证文献
引用本文

唐奕明,刘崇茹.新型电力系统中数字孪生辅助的网络切片资源分配[J].半导体光电,2026,47(2):407-416. TANG Yiming, LIU Chongru. Software-Defined Network Slicing Test Bench and Validation for White-Box Switches[J].,2026,47(2):407-416.

复制
分享
相关视频

文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2025-02-26
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2026-05-25
  • 出版日期:
文章二维码