On the source switching problem of Peer-to-Peer streaming

Zhenhua Li, Jiannong Cao, Guihai Chen, Yan Liu

Research output: Journal article publicationJournal articleAcademic researchpeer-review

7 Citations (Scopus)


Peer-to-Peer(P2P) streaming has been proved a popular and efficient paradigm of Internet media streaming. In some applications, such as an Internet video distance education system, there are multiple media sources which work alternately. A fundamental problem in designing such kind of P2P streaming system is how to achieve fast source switching so that the startup delay of the new source can be minimized. In this paper, we propose an efficient solution to this problem. We model the source switch process, formulate it into an optimization problem and derive its theoretical optimal solution. Then we propose a practical greedy algorithm, named fast source switch algorithm, which approximates the optimal solution by properly interleaving the data delivery of different media sources. The algorithm can adapt to the dynamics and heterogeneity of real Internet environments. We have carried out extensive simulations on various real-trace P2P overlay topologies to demonstrate the effectiveness of our model and algorithm. The simulation results show that our proposed algorithm outperforms the normal source switch algorithm by reducing the source switch time by 20%-30% without bringing extra communication overhead. The reduction in source switching time is more obvious as the network scale increases.
Original languageEnglish
Pages (from-to)537-546
Number of pages10
JournalJournal of Parallel and Distributed Computing
Issue number5
Publication statusPublished - 1 May 2010


  • Multimedia streaming
  • Peer-to-Peer
  • Source switching

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Software
  • Hardware and Architecture
  • Computer Networks and Communications
  • Artificial Intelligence


Dive into the research topics of 'On the source switching problem of Peer-to-Peer streaming'. Together they form a unique fingerprint.

Cite this