Abstract | ||
---|---|---|
A fundamental problem in peer-to-peer streaming is how to select peers from a large network to request their media data. Due to the heterogeneity and the time-varying features of shared resources between peers, an adaptive method is required to select suitable peers. In this paper, we use Hidden Markov Models (HMMs) to model each peer to reflect the variation of resources. Among peers with different HMMs, the one which produces the maximum observation probability is selected as the serving peer. Through simulation results, we show that the proposed algorithm can achieve a good streaming quality and low communication overhead. In addition to these characteristics, the proposed model also comes with the fairness property. |
Year | DOI | Venue |
---|---|---|
2007 | 10.1007/978-3-540-77090-9_13 | EUC Workshops |
Keywords | Field | DocType |
hidden markov models,fairness property,adaptive method,different hmms,hidden markov model,suitable peer,low communication overhead,fundamental problem,proposed algorithm,large network,hmm | Peer-to-peer,Computer science,Adaptive method,Computer network,Maximum observation,Real-time computing,Hidden Markov model | Conference |
Volume | ISSN | ISBN |
4809 | 0302-9743 | 3-540-77089-5 |
Citations | PageRank | References |
0 | 0.34 | 9 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Sheng-De Wang | 1 | 720 | 68.13 |
Zheng-Yi Huang | 2 | 1 | 0.69 |