{"title":"云网络中高效的VNF部署方案","authors":"Jianguo Fu, Guanglei Li","doi":"10.1109/ICCSN.2019.8905349","DOIUrl":null,"url":null,"abstract":"With maturity of cloud network technologies and rapid development of NFV (Network Function Virtualization), the research on NFV based on cloud technologies has gradually increased in academics and industry, aiming to improve scalability, flexibility, and cost-efficiency of network service provision systems. A key challenge of NFV is how to design a cost-effective VNF (Virtualized Network Function) deployment scheme. Most of the existing schemes deploy VNF based on resources requested by users and ignore optimizations during the running time, which causes low capacity utilization. Therefore, we propose an efficient VNF deployment scheme in this paper, which has two parts: over-deployment and automatic migration to improve resource utilization rate and guarantee on-demand capacity provisioning at the same time. The over-deployment approach takes advantage of the characteristic that resources actually used by VNFs are mostly less than the resources requested and deploys VNF based on running load. However, over-deployment may cause PMs (Physical Machines) overloaded at some time, thus, we use the automatic migration approach to solve the overload problem by migrating VNFs. The automatic migration approach moves proper VNFs to lighten the load of overloaded PMs. We perform extensive simulations and numerical results verify benefits of the proposed scheme in terms of the average number of deployed VNFs, active PMs, the rejection rate of VNF requests, and VNF migrations.","PeriodicalId":330766,"journal":{"name":"2019 IEEE 11th International Conference on Communication Software and Networks (ICCSN)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"7","resultStr":"{\"title\":\"An Efficient VNF Deployment Scheme for Cloud Networks\",\"authors\":\"Jianguo Fu, Guanglei Li\",\"doi\":\"10.1109/ICCSN.2019.8905349\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"With maturity of cloud network technologies and rapid development of NFV (Network Function Virtualization), the research on NFV based on cloud technologies has gradually increased in academics and industry, aiming to improve scalability, flexibility, and cost-efficiency of network service provision systems. A key challenge of NFV is how to design a cost-effective VNF (Virtualized Network Function) deployment scheme. Most of the existing schemes deploy VNF based on resources requested by users and ignore optimizations during the running time, which causes low capacity utilization. Therefore, we propose an efficient VNF deployment scheme in this paper, which has two parts: over-deployment and automatic migration to improve resource utilization rate and guarantee on-demand capacity provisioning at the same time. The over-deployment approach takes advantage of the characteristic that resources actually used by VNFs are mostly less than the resources requested and deploys VNF based on running load. However, over-deployment may cause PMs (Physical Machines) overloaded at some time, thus, we use the automatic migration approach to solve the overload problem by migrating VNFs. The automatic migration approach moves proper VNFs to lighten the load of overloaded PMs. We perform extensive simulations and numerical results verify benefits of the proposed scheme in terms of the average number of deployed VNFs, active PMs, the rejection rate of VNF requests, and VNF migrations.\",\"PeriodicalId\":330766,\"journal\":{\"name\":\"2019 IEEE 11th International Conference on Communication Software and Networks (ICCSN)\",\"volume\":\"6 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2019-06-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"7\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2019 IEEE 11th International Conference on Communication Software and Networks (ICCSN)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICCSN.2019.8905349\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 IEEE 11th International Conference on Communication Software and Networks (ICCSN)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCSN.2019.8905349","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 7
摘要
随着云网络技术的成熟和NFV (network Function Virtualization,网络功能虚拟化)的快速发展,学术界和业界对基于云技术的NFV的研究逐渐增多,旨在提高网络服务提供系统的可扩展性、灵活性和成本效益。如何设计一种经济高效的虚拟化网络功能(VNF)部署方案是NFV面临的一个关键挑战。现有的方案大多是根据用户的资源需求部署VNF,忽略了运行时的优化,导致容量利用率较低。因此,本文提出了一种高效的VNF部署方案,该方案分为过度部署和自动迁移两部分,以提高资源利用率,同时保证按需提供容量。过度部署方法利用了VNF实际使用的资源大多少于请求的资源这一特点,并根据运行负载部署VNF。但是,过度部署可能会在某些时候导致pm(物理机)过载,因此,我们使用自动迁移方法通过迁移VNFs来解决过载问题。自动迁移方法移动适当的VNFs,以减轻过载pm的负载。我们进行了大量的模拟和数值结果,验证了所提出方案在部署VNF的平均数量、活动pm、VNF请求的拒绝率和VNF迁移方面的优势。
An Efficient VNF Deployment Scheme for Cloud Networks
With maturity of cloud network technologies and rapid development of NFV (Network Function Virtualization), the research on NFV based on cloud technologies has gradually increased in academics and industry, aiming to improve scalability, flexibility, and cost-efficiency of network service provision systems. A key challenge of NFV is how to design a cost-effective VNF (Virtualized Network Function) deployment scheme. Most of the existing schemes deploy VNF based on resources requested by users and ignore optimizations during the running time, which causes low capacity utilization. Therefore, we propose an efficient VNF deployment scheme in this paper, which has two parts: over-deployment and automatic migration to improve resource utilization rate and guarantee on-demand capacity provisioning at the same time. The over-deployment approach takes advantage of the characteristic that resources actually used by VNFs are mostly less than the resources requested and deploys VNF based on running load. However, over-deployment may cause PMs (Physical Machines) overloaded at some time, thus, we use the automatic migration approach to solve the overload problem by migrating VNFs. The automatic migration approach moves proper VNFs to lighten the load of overloaded PMs. We perform extensive simulations and numerical results verify benefits of the proposed scheme in terms of the average number of deployed VNFs, active PMs, the rejection rate of VNF requests, and VNF migrations.