Proxy-Based Peer-to-Peer Network: Analysis, Optimization and Algorithms

Peer-to-peer (P2P) technique has emerged as a promising scalable solution for live streaming to large group. In this talk, we address the design of a proxy-based peer-to-peer network, where proxies are used as “backbone” to carry streams to local users (peers). Users, after mapping to their nearby proxies, get their streams in a peer-to-peer manner. Our goals are to achieve low source-to-peer delay, high stream quality, and robustness against churns.

To achieve that in the presence of diverse uplink bandwidth, a natural choice is the use of mesh for both proxy and peer networks, where each proxy or peer is served by multiple parents. Since user delay is the longest path delay from the source, it increases quickly with the size and depth of the mesh. The problem is then how to design a low-delay mesh so as to meet a certain streaming rate requirement.

We present the stochastic analysis, problem formulation and optimization of streaming mesh. The optimization problem is shown to be NP-hard. We then present distributed algorithms which continuously and dynamically improve user delay. Through simulation and PlanetLab experiments, we show that our distributed algorithms substantially outperform traditional and state-of-the-art approaches, and achieve close to optimal delay based on complete knowledge. There is also an optimal number of parents for a peer to achieve the lowest delay in the mesh.

Speaker Details

S.-H. Gary Chan (S’89-M’98-SM’03) received MSE and PhD degrees in Electrical Engineering from Stanford University (Stanford, CA) in 1994 and 1999, respectively, with a minor in business administration. He obtained his B.S.E. degree (highest honor) in Electrical Engineering from Princeton University (Princeton, NJ) in 1993, with certificates in applied and computational mathematics, engineering physics, and engineering and management systems. Multimedia Tools and Applications (2007). He was a co-chair of multimedia symposium in IEEE Globecom (2007 and 2006) and IEEE ICC (2007 and 2005), and for the workshop on “Advances in Peer-to-Peer Multimedia Streaming” in ACM Multimedia Conference (2005).Dr. Chan is a member of Tau Beta Pi, Sigma Xi, and Phi Beta Kappa. He was a William and Leila Fellow at Stanford University in 1993-94. At Princeton, he was the recipient of the Charles Ira Young Memorial Tablet and Medal, and the POEM Newport Award of Excellence in 1993. His research interest includes multimedia networking, peer-to-peer technologies and streaming, and wireless communication networks.He is currently an Associate Professor with the Department of Computer Science and Engineering, The Hong Kong University of Science and Technology (HKUST), Hong Kong. He is also an Adjunct Researcher with Microsoft Research Asia in Beijing. He served as the Director of Computer Engineering Program at the HKUST from 2006 to 2008. He was a Visiting Assistant Professor in networking at the Department of Computer Science, University of California at Davis, CA, from 1998 to 1999. During 1992-93, he was a Research Intern at the NEC Research Institute, Princeton, NJ.Dr. Chan is currently an Associate Editor of IEEE Transactions on Multimedia, and a Vice-Chair of Peer-to-Peer Networking and Communications Technical Sub-Committee, IEEE Comsoc Emerging Technologies Committee. He is the TPC chair of IEEE Consumer Communications and Networking Conference (CCNC) in 2010. He served as a Vice-Chair of IEEE COMSOC Multimedia Communications Technical Committee (MMTC) from 2003 to 2006. He was a guest editor of the special issue on “Peer-to-Peer Multimedia Streaming” in IEEE Communication Magazine (2007) and “Advances in Consumer Communications and Networking” in Springer.

Date:
Speakers:
S. H. Gary Chan
Affiliation:
Hong Kong University of Science and Technology