The pull of Software Defined Networking (SDN) which promises the formidable configuration, vigorous evolution and satisfactory performance is appealing. However, SDN is still in its initial period. Many issues remain to be delved and the most intractable one is the scalability of control plane with the explosive increase of network scales. In order to address this issue, splendid researches have presented multiple controllers to implement logically centralized control layer. Whilst, with the growing number of controllers, another hot potato emerges, which is the energy consumption. In our previous research, we proposed multiple controllers sleeping management which reduces the energy but introduces more delay. In this paper, based on the previous research, we propose the energy saving solution based on big data analysis. We utilize big data to analyse topology data in the system and excavate influence of each controller according to the analytic results. On this basis, the system opts to switch off controllers with smaller influence so as to reduce delay. Simulation results reveal that the improved method with big data analysis achieves superior time delay compared with the previous one.

Additional Metadata
Keywords Big data, Energy consumption, Green SDN, Sleeping management
Persistent URL
Conference 2016 IEEE Globecom Workshops, GC Wkshps 2016
Qiu, C. (Chao), Zhao, C. (Chenglin), Yao, H. (Haipeng), Xu, F. (Fangmin), & Yu, F.R. (2016). Why did you opt to switch off me? Big data for green Software Defined Networking. In 2016 IEEE Globecom Workshops, GC Wkshps 2016 - Proceedings. doi:10.1109/GLOCOMW.2016.7849083