Research Article
Approximation towards energy-efficient distributed environments
@INPROCEEDINGS{10.4108/ICST.SIMUTOOLS2010.8677, author={Alberto E. Garcia and Roman Weidlich and Laura Rodriguez de Lope and Klaus D. Hackbarth and Helmut Hlavacs and Caridad San Leandro}, title={Approximation towards energy-efficient distributed environments}, proceedings={The industry track of SIMUTools 2010.}, publisher={ACM}, proceedings_a={SIMULATIONWORKS}, year={2010}, month={5}, keywords={Virtual Home Environment cost model energy efficiency P2P virtualization}, doi={10.4108/ICST.SIMUTOOLS2010.8677} }
- Alberto E. Garcia
Roman Weidlich
Laura Rodriguez de Lope
Klaus D. Hackbarth
Helmut Hlavacs
Caridad San Leandro
Year: 2010
Approximation towards energy-efficient distributed environments
SIMULATIONWORKS
ICST
DOI: 10.4108/ICST.SIMUTOOLS2010.8677
Abstract
Peer-To-Peer (P2P) traffic represents almost 60 % of Internet traffic, and involves personal computers working during periods of full-time operation (Always-ON). The rational use of shared resources opens up the possibility of reducing energy consumption associated with this type of operation. This paper analyses the results previously obtained from simulating an environment that shares resources among multimedia home networks, called Virtual Home Environment (VHE) and sponsored by the European Network of Excellence EuroFGI [1]. The study analyzes the behavior of basic P2P distributed environments, comparing them to the new environment based on the virtualization of specific processes. This proposal includes a cost model that enables the restriction of behaviors associated with hoarding of resources. As a result, the energy consumption associated with this improvement implies a substantial reduction in the number of Always-ON devices, and the reduction and equalization of activity time around the area of the distributed network. The proposed simulation will be applied in future developments related with the activities proposed by Cost Action IC0804: Energy efficiency in large-scale distributed systems, see [2].