Skip Navigation

Home / Graduate Conference

2012 Master's Level Graduate Research Conference

Session I - Session II - Session III - Session IV

Performance Evaluation of Multi-domain GENI Networks

This presentation identifies the comprehensive study of a GENI network through the performance evaluation of such a network. The IP tos fields can be modified to measure the Quality of Service(QoS) parameters. Depending on the parameters, the high priority traffic will be given preference over the low priority traffic in the network. We are testing the QoS parameters through our implementation and study the behavior of high priority traffic using the GENI resources. The project aims at making use of the resources available on the test bed using them for analyzing the performance metrics. Any communication network supports multitude of data and applications such as delay sensitive real time voice or high definition video streaming. Bandwidth-intensive applications stretch network capacity, so it is essential to define the metrics of secure predictable and guaranteed services. The Quality of Service can be achieved by managing the factors such as jitters, delay, delay variation and packet loss parameters. There are currently two different approaches to IP packet switching networks, which include Integrated Services and differentiated services for QoS. On the GENI test bed network, we will develop techniques to allocate network resources using Openflow switches. The Openflow switches will be configured along with the controller and flow visor for forwarding the traffic. Two nodes, GUI based virtual machines will be allocated from two different geographical locations. Each of the client nodes will be able to stream the video and voice from the server using the VideoLAN application (VLC media player). The switches will be configured accordingly with respect to QoS parameters so that they can send the video and voice traffic. 2. Justification: We are working on this idea that includes the integration of QoS testing using GENI resources. This project will help analyze and understand the Openflow switch configurations along with details of the video streaming over resources reserved on the GENI test bed. This will further help to understand the difference in behavior between defined level of quality of service and observed metrics. By testing the priority parameter, this further helps understand and experiment on QoS factors using the experimental slice and sliver as expected by the requirements of the project.

Presenters: Anand Akella (Rochester Institute of Technology) --
Praveen Iyengar Parathasarthy (Rochester Institute of Technology) --
Topic: Math and Sciences - Poster Session
Location: Edwards Hall Lobby
Time: 9 am (Session I)