|
|
|
|
|
OPNET Technologies 7255 Woodmont Avenue Bethesda, MD 20814 |
||
|
Tel:
240-497-3000 Fax: 240-497-3001 E-mail: university@opnet.com Web: www.opnet.com |
|
Project Title |
On Demand Media Streaming Over Heterogeneous Peer-to-Peer Networks |
|
Researchers |
Dr. Nabil J. Sarhan,
Assistant Professor,
nabil@ece.eng.wayne.edu
Bashar Qudah, PhD Candidate: bqudah@wayne.edu Iyad Fayez Jafar, PhD Student, iyad@wayne.edu |
|
Department |
Electrical and Computer Engineering |
|
Why We Need OPNET? |
To evaluate new techniques for P2P streaming of multimedia contents over the Internet |
|
Project Description |
Streaming multimedia over the Internet has recently captured the interest of many research studies due to the wide and rapid deployment of high speed networks like DSL and Cable modems. Numerous applications make use of such advancements. Examples include distance learning, VoD and digital libraries. Client-server model used in the WWW service was the conventional way to provide such services. Nevertheless, and because of the nature of media data, this model does not scale well to adapt for the excessive demand on the service. Several schemes were proposed to boost the performance of the traditional model. The common idea among those schemes is to make media streams sharable by utilizing the IP multicast service. With those methods in hand, we still need to solve the issues and obstacles of multicast deployment in the Internet. As a substitute, peer-to-peer (P2P) networks were targeted in the past few years. In such systems, peers (clients) communicate with other peers in a server-less environment to share data as well as other resources such storage and CPU. Peers involved in the communication usually do not posses server like behavior. On the other hand, they are heterogeneous and characterized by low bandwidth, availability and storage capacity. In this project, we will address the heterogeneity problem in P2P networks. Mainly, we are concerned on how the limited storage space offered by the peers will affect the system performance. Thus, different dispersion algorithms, admission control policies, and fault tolerance mechanisms will be designed in way to take into account this restriction all together with other factors like the offered bandwidth, and peers' availability. What follows is the implementation using OPNET to evaluate the system performance in terms of admission rate, QoS, and service latency.
|