| 摘要: |
| 在移动边缘计算的动态网络服务系统中，综合考虑用户请求流的优先级、混合服务功能链（Hybrid Service Function Chain,H-SFC）的顺序和延迟约束、虚拟化网络功能（Virtual Network Functions,VNF）的共享以及用户的服务时间，研究H-SFC的在线部署问题。首先，通过整数线性规划对该问题建模。然后，以最大化网络的累计请求流接受率为目标，提出一种基于优先级的H-SFC在线部署策略。该策略由可用资源的预估算法（Estimation of Available Resources Algorithm，EARA）和基于优先级的部署算法（Priority-based Deployment Algorithm，PBDA）两部分组成。EARA算法采用整合移动平均自回归模型进行资源的快速预测，然后通过调整和释放已离开的移动用户所占用的资源来为PBDA提供资源依据。PBDA根据EARA所预测的可用资源情况，使用优先队列的记忆化深度优先搜索方法为H-SFC选取所需的VNF节点和链路。仿真结果表明，该策略的流接受率比不用EARA算法调整时提高了10%。另外，在其他场景下也展现出更好的性能。 |
| 关键词: 移动边缘计算 混合服务功能链 在线部署 |
| DOI：10.20079/j.issn.1001-893x.240512001 |
|
| 基金项目:重庆市教育科学规划课题（K22YG224267）；山西省基础研究计划资助项目(202403021212167)；山西省高等学校科技创新项目（Z48019） |
|
| Priority-based Hybrid Service Function Chain Online Deployment Strategy |
| WANG Liyan,ZHENG Xianfeng,HOU Zetian,FENG Hao |
| (1.Big Data College,Chongqing College of Mobile Communication,Chongqing 401420,China;2.Chongqing Key Laboratory of Public Big Data Security Technology,Chongqing 401420,China;3.Bishan Branch,China Mobile Chongqing Co.,Ltd.,Chongqing 402760,China;4.School of Information,Shanxi University of Finance and Economics,Taiyuan 030006,China) |
| Abstract: |
| The online deployment strategy of hybrid service function chains(H-SFCs) in dynamic network service systems at the mobile edge computing is investigated,which jointly considers the priority of user request flows,sequential constraints of H-SFC,delay constraints,resource sharing of virtual network functions(VNFs),and the service time of user.Firstly,the problem is formulated as integer linear programming.Then,a priority-based online deployment strategy is proposed to maximize the cumulative acceptance rate of network flows during the monitoring period.The strategy consists of two parts,the estimation of available resources algorithm(EARA) and the priority-based deployment algorithm(PBDA).The EARA algorithm uses an autoregessive integrated moving average model for rapid resource prediction,and adjusts the resource allocation of VNFs and releases resources occupied by departed mobile users to obtain available resources for PBDA.PBDA uses the memorized depth first search based on priority queue to select the required VNF nodes and links for H-SFC based on the available resource information provided by EARA.The simulation results show that the flow acceptance rate of this strategy is improved by 10% compared with that of the EARA algorithm being not used for adjustment.In addition,the strategy also shows better performance in other scenarios. |
| Key words: mobile edge computing hybrid service function chain online deployment |
《电讯技术》编辑部 