面向动态边缘网络的数字孪生自适应联合优化部署  

Adaptive joint optimization deployment of digital twin for dynamic edge networks

作　　者：张晖[1] 罗天翔 王倩倩[2] ZHANG Hui;LUO Tianxiang;WANG Qianqian(Internet of Things Research Institute,Nanjing University of Posts and Telecommunications,Nanjing 210003,China;School of Software Engineering,Jinling Institute of Technology,Nanjing 211169,China)

机构地区：[1]南京邮电大学物联网研究院,江苏南京210003 [2]金陵科技学院软件工程学院,江苏南京211169

出　　处：《南京邮电大学学报（自然科学版）》2024年第2期1-10,共10页Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition

基　　金：国家重点研发计划(2020YFB2104004);国家自然科学基金(62071252);江苏省重点研发计划(BE2021725);江苏省基础研究计划(自然科学基金)前沿引领技术基础研究专项(BK20212001);江苏省高校自然科学研究重大项目(21KJA510005)资助项目。

摘　　要：在数字孪生赋能的移动边缘计算(Mobile Edge Computing,MEC)网络中,如何实现数字孪生服务器的高效部署是确保数字孪生实时交互性的瓶颈问题。针对该问题,提出一种面向动态边缘网络的数字孪生自适应联合优化部署机制。首先,该机制构建面向动态边缘网络的双层数字孪生模型以实时捕捉MEC网络状态和UE资源利用情况等特征。然后,联合数字孪生交互时延模型、负载均衡模型和能源消耗模型建立数字孪生服务器自适应动态更新部署问题。最后,提出多阶段自适应动态联合部署优化算法,将数字孪生服务器自适应动态更新部署问题分解为数字孪生服务器初始化部署和自适应动态更新部署两阶段优化求解,以实现部署策略随MEC网络的即时系统状态进行自适应动态调整。仿真分析验证了所提出算法在预测精度、交互时延、工作负载和能耗方面的有效性。In digital twin-enabled Mobile Edge Computing(MEC)networks,how to realize the efficient deployment of digital twin servers is the bottleneck problem to ensure the real-time interactivity of digital twin.To address this problem,an adaptive joint deployment optimization mechanism of digital twin servers for dynamic edge networks is proposed.First,the mechanism constructs a two-tier digital twin model for dynamic edge networks to capture features such as MEC network status and UE(user equipment)resource utilization in real-time.Secondly,the digital twin server adaptive dynamic update deployment problem is established by jointly using the digital twin interaction latency model,the load balancing model,and the energy consumption model.Finally,a multi-stage adaptive joint deployment optimization algorithm is proposed to decompose the digital twin server adaptive dynamic update deployment problem into two stages for optimization and solution,including initial deployment for the digital twin server and the adaptive dynamic update deployment for the digital twin server.Thus,the adaptive dynamic adjustment of the deployment policies with the immediate system status of the MEC networks is established.Simulation analysis verifies the effectiveness of the proposed multi-stage adaptive joint deployment optimization algorithm in terms of prediction accuracy,interaction latency,workload,and energy consumption.

关 键 词：数字孪生网络 服务器部署 多目标优化 动态边缘网络 物联网系统 

分 类 号：TN929.5[电子电信—通信与信息系统]