دانلود رایگان مقاله انگلیسی معماری SDN و NFV مبتنی بر ابر برای زیرساخت های اینترنت اشیا - الزویر 2018

abstract

Connected devices which are commonly known as Internet of Things (IoT) are increasing at an alarming pace. Network infrastructure has to accommodate all these devices by providing adequate connectivity and delivering application based services. Service providers have to invest more in network infrastructure to meet the growing needs. To address this issue, a new concept of infrastructure sharing among service providers has emerged to reduce excessive investment costs related to infrastructure deployment. For this reason new architectures based on network virtualization are emerging to provide network sharing, handle Big data explosion from IoT devices, and simplify management tasks. Two complementary architectures, software defined networking (SDN) and network function virtualization (NFV) are emerging to comprehensively address several networking issues. In this work, we introduce the most embraced virtualization concepts proposed by SDN and NFV architectures. We quantitatively evaluate hardware and energy cost savings with these two SDN and NFV architectures compared to the existing state-of-the-art network 4G hardware technologies.

7. Conclusion

The future network challenges and the role of virtualization in addressing all issues related to cloud based SDN and NFV technologies were discussed. Further, the concepts of virtualization were discussed along with the concepts from SDN and NFV frameworks. The necessity of cloud based data center networks is discussed from the view point of big data explosion and how the SDN/NFV based technology will likely handle the big data explosion over cloud networks. Furthermore, a cloud based SDN/NFV network was studied and a mathematical model is presented that compares the cost and energy consumption between the SDN/NFV network and a typical 4G network. All key metrics are taken as variable functions to study their effect on the overall cost and energy consumption in the SDN/NFV network. Adhering to the common assumptions in the literature, the proposed model investigates the relative cost and energy consumption for both single service provider and all the service providers in the system as a whole that are involved in SDN/NFV network sharing. By eliminating the possibility of any multiplexing gain, we have still found considerable cost reductions and energy savings in SDN/NFV based networks. The results have substantiated the claims of many gains achievable tualization. In our future work, we are interested in seeing how the opportunistic resource access and the multiplexing gains will provide more energy and cost savings in the shared SDN/NFV network.