Javascript must be enabled for the correct page display

Context Aware Power Management based on User Behaviour

Setz, B. (2015) Context Aware Power Management based on User Behaviour. Master's Thesis / Essay, Computing Science.

[img]
Preview
Text
MasterThesisBrianSetzFinal.pdf - Published Version

Download (2MB) | Preview
[img] Text
toestemming.pdf - Other
Restricted to Backend only

Download (615kB)

Abstract

Personal computers are responsible for a significant portion of the energy consumption of office buildings. There is little incentive for occupants of these types of buildings to save energy, as they are not responsible for paying the electricity bills. Enabling power management options such as sleep mode is a low effort method to reduce the energy consumption of computers. However, choosing the right timeout before a computer goes to sleep can be challenging. A sleep timeout which is too small leads to discomfort, whereas a sleep timeout which is too large results in poor energy saving efficiency. Furthermore, each user has his or her own preference when it comes to the sleep timeout. Letting the user choose their own sleep timeout is not an option, since research shows most users disable the sleep timeout completely. In this thesis we attempt to find a solution which can determine the sleep timeout based on user behavior. Multiple models have been designed with the goal of maximizing the energy savings while minimizing discomfort. Each model has a slightly different approach to learning the optimal sleep timeout. A software solution was implemented to collect the data required to use and analyze the models. As part of this thesis we shall take a look at the architecture of this software solution and the supporting infrastructure of services which allows high performance while remaining scalable. The models have been tested on the computers of employees of the University of Groningen during several weeks. We will analyze the results of the experiments and determine which model performs best according to some predefined criteria. We shall also take a look at the energy savings and economical savings which could be achieved if this solution were to be deployed on every computer in the Bernoulliborg.

Item Type: Thesis (Master's Thesis / Essay)
Degree programme: Computing Science
Thesis type: Master's Thesis / Essay
Language: English
Date Deposited: 15 Feb 2018 08:08
Last Modified: 15 Feb 2018 08:08
URI: http://fse.studenttheses.ub.rug.nl/id/eprint/13342

Actions (login required)

View Item View Item