ManSDN/NFV Keynote Talks
Title: Energy Efficiency in Cloud Networks
Speaker: Prof. Jaafar Elmirghani, University of Leeds, UK.
Date: November 4th, 2016, 09:00 - 10:00 AM
Bio: Prof. Jaafar Elmirghani (PI) is FIET, FIoP, and Director of the Institute of Integrated Information Systems, Leeds. He joined Leeds in 2007 and prior to that founded, developed and directed the Institute of Advanced Telecommunications (ERDF £5.2m), and the Technium Digital (ERDF £4.3m) at Swansea University. He has provided outstanding leadership in a number of large research projects, secured over £22 million in grants over the past 8 years and is currently PI of the £6m EPSRC INTERNET Programme Grant (2010-2016) with Cambridge. He led between 2010 and 2015 the Wired Core and Access Networks (WCAN) working group of GreenTouch, one of two technical committees in GreenTouch representing about half of the 50+ GreenTouchindustrial and academic member organisations. He was awarded in international competition the IEEE Comsoc 2005 Hal Sobolaward, the IEEE Comsoc outstanding service award in 2009 for “contributions to signal processing and communication electronics”, and in 2015 for “leadership and contributions to the area of green communications”, the 2015 GreenTouch 1000x award for “pioneering research contributions to the field of energy efficiency in telecommunications” he shared the 2016 Edison Award in the collective disruption category with a team of 6 from GreenTouch for their joint work on the GreenMeter and was awarded the 2016 Premium Award for best paper in IET Optoelectronics for work on Green Optical OFDM networks. He has published over 450 technical papers, and has research interests in communication networks and systems.
Abstract: Cloud computing is expected to be a major factor that will dominate the future Internet service model. This talk summarizes our work on energy efficiency for cloud networks. We describe a framework for studying the energy efficiency of four cloud services in IP over WDM networks: cloud content delivery, storage as a service (StaaS), and virtual machines (VMS) placement for processing applications and infrastructure as a service (IaaS). Our approach is based on the co-optimization of both external network related factors such as whether to geographically centralize or distribute the clouds, the influence of users’ demand distribution, content popularity, access frequency and renewable energy availability and internal capability factors such as the number of servers, switches and routers as well as the amount of storage demanded in each cloud. Our investigation of the different energy efficient approaches uses Mixed Integer Linear Programming (MILP) models and real time heuristics.