Green IT - dynamic network topologies

Costantini, Daniel (2015) Green IT - dynamic network topologies. [USQ Project]

[img]
Preview
Text
Costantini_D_Kist.pdf

Download (1655Kb) | Preview

Abstract

All engineering disciplines are influenced by the global focus on energy consumption reduction and sustainability. Due to its resident inefficiency, The ICT sector is of particular concern, and there has been extensive work to develop sustainability enhancements to networks and/or network devices. Previous work presented dynamic topology concepts in which the behaviour and topology of the devices and the network react dynamically in response to traffic demands, with the intent of placing devices into standby states to reduce energy consumption. The key aim of this
study is to develop a dynamic topology mechanism implementation; it proposes a testbed environment
and corresponding dynamic topology mechanism that makes use of two programs: one running on a centralised controller, and one running on the network nodes. The former determines the optimal topology based on energy consumption reductions and network traffic, while the latter uses MPLS to implement the topology. The testbed is used to determine the dynamic topology mechanism’s effectiveness and impact on network performance, and does so by subjecting it to controlled variations in network traffic. Quantitative measurements of the dynamic topology mechanism’s network performance metrics are presented and analysed relative to baseline measurements. The analysis shows that the dynamic topology mechanism is quite effective, as the effect on network performance is mostly minimal and the reaction to network traffic variations is sufficiently swift. The system takes approximately 30 seconds to react to traffic variations and implement topology changes, and has negligible effect on jitter, packet loss, and the number of out of order packets. However, it produces an average increase in delay of 8 ms, the source of which requires further investigation. This study proves the feasibility of dynamic topology mechanism implementation, and provides a framework for further development and eventual widespread deployment.


Statistics for USQ ePrint 29190
Statistics for this ePrint Item
Item Type: USQ Project
Item Status: Live Archive
Additional Information: Bachelor of Computer Systems Engineering (Honours) project
Faculty / Department / School: Current - Faculty of Health, Engineering and Sciences - School of Mechanical and Electrical Engineering
Supervisors: Kist, Alexander
Date Deposited: 01 Jun 2016 01:28
Last Modified: 06 Jun 2016 01:08
Uncontrolled Keywords: Green IT, Dynamic networks, Linux, MPLS, Raspberry Pi
Fields of Research : 09 Engineering > 0906 Electrical and Electronic Engineering > 090602 Control Systems, Robotics and Automation
URI: http://eprints.usq.edu.au/id/eprint/29190

Actions (login required)

View Item Archive Repository Staff Only