arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2007.09789cs.NI
收藏

支持vSDN的5G网络中的自适应控制平面负载均衡

Adaptive Control Plane Load Balancing in vSDN Enabled 5G Network

Deborsi Basu, Uttam Ghosh, Raja Datta

摘要:

在这项工作中,我们制定了一个集群管理程序(C-H)对部署方法,以提供各种网络功能(如控制和数据信号之间的差异,不同的翻译功能等)。低延迟。该系统模型处理一个真实的网络拓扑结构和四个定义良好的网络延迟矩阵与MILP模型已被很好地构造,以优化延迟目标。反向路径流机制(RPFM)的提出是为了提供可行的解决方案,保持网络负载和链路容量的约束下的容忍限制。我们进一步减少了H-平面负载分配的网络资源的基础上的服务在请求的到达时间从用户。通过广泛的实验,所提出的系统的功效已被证明在显着的延迟减少和e-ESTA(进化超低延迟)的经验,有效地处理所有实时用户的需求。所提出的方法可用于类似的定位问题,如5G-NR中的服务链映射,5G C-RAN中的基带单元部署,分布式CPS中的防火墙部署。

英文摘要:

In this work, we have formulated a controllerhypervisor (C-H) pair deployment approach to provide various network functions (such as differentiation between control and data signals, different translation functions etc.) with low latency. The system model deals with a real network topology and four well defined network latency matrices with an MILP model has been well constructed to optimize latency objectives. Reverse Path-Flow Mechanism (RPFM) has been proposed in order to provide feasible solutions by keeping the network load and link capacity constrained under tolerance limit. We further minimize the H-plane load by distributing the network resources based on the arrival time of SERVICE IN requests from users. Through extensive experiment, the efficacy of the proposed system has been demonstrated in terms of significant latency reduction and an e-ULL (evolved-Ultra Low Latency) experience with efficient handling of all real-time user demands. The proposed approach can be used for similar localization problems like service chain mapping in 5G-NR, baseband unit deployment in 5G C-RAN, firewall deployment in distributed CPS.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑