Back to Search
Start Over
Quality enhancement of video on demand implementation in peer-to-peer networks by optimum chunk length in the BitTorrent.
- Source :
- Peer-to-Peer Networking & Applications; Jul2017, Vol. 10 Issue 4, p1051-1062, 12p
- Publication Year :
- 2017
-
Abstract
- BitTorrent in efficiency content distribution is a major concern efficiency among the researchers of this field with respect to streaming video on demand (VoD) production. BitTorrent is not appropriate for real-time applications; therefore, in order to apply it in VoD it should go through the necessary changes. Most of the available studies have greatly focused on changes in methods regarding chunk and peer selections method regarding BitTorrent, which proposed methods have improved the quality of VoD to a certain degree, while, the effect of chunk size on quality of video has been of less concern among them. Noting the fact that the buffer is used on VoD, the specified time for filling the buffer would allow for appropriate management of the chunk's length. The Bit error rate and the time overhead of the operating algorithm parameters, somehow effect the chunk size. Because of bit error rate, the probability of correctly received chunks with great length is much less, that is, offering shorter pieces, while these pieces would lead to formation of more pieces in a buffer. The results indicate that a specific amount of time is required for obtaining the buffer's content, and it must be dividable into more chunks. Running algorithms for each chunk generate a greater overhead which would result in of the QoS reduction. This overhead would make the bigger pieces perform better. As for the opposite impacts of these two parameters on the chunk size, in this article the optimal length of the chunks is found by considering both the effective characteristic. This optimal length is an established balance between the correctly received chunks' rate and the greater rate of the buffer context obtained in a specified time. [ABSTRACT FROM AUTHOR]
Details
- Language :
- English
- ISSN :
- 19366442
- Volume :
- 10
- Issue :
- 4
- Database :
- Complementary Index
- Journal :
- Peer-to-Peer Networking & Applications
- Publication Type :
- Academic Journal
- Accession number :
- 122761416
- Full Text :
- https://doi.org/10.1007/s12083-016-0467-y