TY - GEN
T1 - Reducing energy consumption of data centers using optical wireless links
AU - Rosenkrantz, Etai
AU - Arnon, Shlomi
N1 - Publisher Copyright: © 2016 IEEE.
PY - 2016/1/1
Y1 - 2016/1/1
N2 - Data Center (DC) servers consume a huge amount of energy. However, a part of this energy (cooling and information technology devices) can be saved by shutting down racks containing idle servers. This can be achieved by rerouting workloads from partially loaded racks, and thus allowing additional racks to switch into idle mode. Workload rerouting requires additional wired links or by optical wireless communication (OWC) links, which can be deployed on top the existing network. In comparison to classic wired infrastructure, OWC deployment and maintenance is faster and simpler, and it offers dynamic configuration of the network. Moreover, OWC can be used to augment the performance of virtualization techniques, which allow load balancing. In this paper we developed a mathematical model and an optimization algorithm along with numerical simulations, which evaluate the energy consumption in DCs. Results show that for a uniform distributed load between 24-100% and a 100% deployment of OWC links, 33% of the IT and cooling energy consumed can be saved.
AB - Data Center (DC) servers consume a huge amount of energy. However, a part of this energy (cooling and information technology devices) can be saved by shutting down racks containing idle servers. This can be achieved by rerouting workloads from partially loaded racks, and thus allowing additional racks to switch into idle mode. Workload rerouting requires additional wired links or by optical wireless communication (OWC) links, which can be deployed on top the existing network. In comparison to classic wired infrastructure, OWC deployment and maintenance is faster and simpler, and it offers dynamic configuration of the network. Moreover, OWC can be used to augment the performance of virtualization techniques, which allow load balancing. In this paper we developed a mathematical model and an optimization algorithm along with numerical simulations, which evaluate the energy consumption in DCs. Results show that for a uniform distributed load between 24-100% and a 100% deployment of OWC links, 33% of the IT and cooling energy consumed can be saved.
KW - Energy consumption
KW - Optical wireless communication
KW - Optimization
UR - http://www.scopus.com/inward/record.url?scp=85017307512&partnerID=8YFLogxK
U2 - 10.1109/WiSEE.2016.7877306
DO - 10.1109/WiSEE.2016.7877306
M3 - Conference contribution
T3 - Proceedings - WiSEE 2016: 2016 IEEE International Conference on Wireless for Space and Extreme Environments
SP - 68
EP - 72
BT - Proceedings - WiSEE 2016
T2 - 2016 IEEE International Conference on Wireless for Space and Extreme Environments, WiSEE 2016
Y2 - 26 September 2016 through 28 September 2016
ER -