Live network multimedia applications (e.g., video conferencing, TV on demand) have been very popular in recent years and are expected to dominate Internet traffic in the near future. With multimedia and Internet-enabled devices being ubiquitous, mechanisms that ensure multimedia flows do not congest the Internet are crucial components of multimedia systems that are embraced rather than opposed by network service providers. The emergence of browser-based multimedia conferencing applications using the WebRTC protocol, an open source project aiming at Real-Time Communication (RTC) with Web, and wide deployment of these applications are expected to increase the traffic of interactive real-time multimedia on the Internet. RTP Media Congestion Avoidance Technique (RMCAT) may be applied to WebRTC, but this is a long-term process and WebRTC deployments will occur before RMCAT is completed. New methods and quick solutions are therefore required to protect the network from uncontrolled media flows until deployment of effective congestion control can be guaranteed. The RTP Protocol Circuit Breaker (RTP-CB) has been proposed in March 2012 within the Internet Engineering Task Force (IETF). Rather than providing congestion control, the RTP-CB is designed only to protect the network by terminating RTP/UDP flows that cause excessive congestion. While the deployment of congestion control for RTP/UDP flow remains an open issue, design a RTP-CB as a quick solution for protecting the current internet is the main focus of this work. In this work by analysing the UDP traffic over a limited path, a RTP-CB algorithm is designed. Then a packet sniffer's code (C routine) is written to sniff and analyse all RTP/UDP, TCP, RTCP SR, and RTCP RR traffic. Based on the designed algorithm the above code was developed further to work as a RTP-CB. This RTP-CB can be deployed on receiver or sender. After deployment of RTP-CB for RTP/UDP flows in a controlled network, its performance in a range of scenarios with using only its congestion rule has been evaluated. The evaluation showed some short coming in performance of RTP-CB in some certain condition when RTP-CB used only congestion rule. The performance of the RTP-CB is evaluated from two perspectives: First, the thesis considered network performance metrics, such as the frequency at which a RTP circuit breaker triggered. Then, it considered the experience of multimedia users, accounting for all outcomes to all users: those congesting the network (where the flow is terminated), those that did not (and are rewarded by reduced congestion) as well as flows that, without severely congesting the network, obtained little quality from a multimedia session and consumed network resources to no avail. Building on the knowledge gathered in these experiments, some extensions (Media Usability Rule) to the RTP-CB rules is proposed and evaluated. This work demonstrates this evaluation by streaming video flows over IP networks using a dedicated test-bed and proposed RTP-CB. These experiments assess the effect of network conditions (packet loss, jitter and network capacity constraint) on the transmission of different types of video stream with and without the proposed RTP-CB Media usability rule. The experiments prove that RTP-CB implementing the congestion rule alone can offer adequate protection to a network, but it does not perform well in some conditions, for example, when the bottleneck buffer size is small. Experiments confirm that the proposed (computationally inexpensive) modifications to the RTP-CB rules improve the RTP-CB performance. The results of these experiments and media usability rule were introduced in IETF RTP-CB draft version 07 of October 27, 2014 and later versions acknowledged contributions by the author of this thesis.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:680972 |
| Date | January 2015 |
| Creators | Fough, Nazila |
| Publisher | University of Aberdeen |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://digitool.abdn.ac.uk:80/webclient/DeliveryManager?pid=228630 |
Page generated in 0.0521 seconds