Sitemap  |  Contact  |  Home  |  中文  |  CAS  |  Director’s Email
International Cooperation
Education & Training
Societies & Publications
Chinese Journal of Acoustics
 
 
  Location:Home>Research>Research Progress
An Adaptive Neighbor Selection Method for P2P Media Streaming System Based on Peer Capacity
Author:
ArticleSource:
Update time: 2010/08/09
Viewed:
Close
Text Size: A A A
Print

 

In practical P2P network, capacity of peer is quite different due to many factors, such as end-to-end delay, upstream bandwidth, memory and so on. In order to improve the performance, choosing neighbors with strong ability is very important in P2P media streaming system. However, at present, most of researches don't pay enough attention to capacity of peer, so FENG Zhentan, NI Hong and WANG Jinlin of Institute of Acoustics, Chinese Academy of Sciences carried out a series of studies and proposed a new method to solve this problem.

Firstly, they propose a model for P2P streaming system. Peers are divided into different levels based on their capacity. Meanwhile, they introduce a description method based on peers' dynamic capacity. Secondly, aiming to the peers' capacity changing with different neighbors, they present an adaptive method called ASDC (Adaptive neighbor Selection method based on peer Dynamic Capacity). ASDC selects the neighbors by random walk method. The key point is setting the expected stationary probability of random walk according to the dynamic peer capacity. The transition probability matrix is calculated through the Metropolis-Hastings methods in order to choose higher level of system nodes with greater probability. Moreover, they elaborate the neighbor's updated method to ensure the node load balancing.

Simulation results demonstrate that the algorithm can significantly improve system performance, reduce system latency, and has a strong robustness in dynamic network environment.

 
Copyright © 1996 - 2020 Institute of Acoustics, Chinese Academy of Sciences
No. 21 North 4th Ring Road, Haidian District, 100190 Beijing, China
E-mail: ioa@mail.ioa.ac.cn