Summary

International Conference on Emerging Technologies for Communications

2020

Session Number:E3

Session:

Number:E3-1

A Study on Improvement of BitTorrent Pieces Acquisition Efficiency by Introducing Collaborative Group of Nearby Peers

Yusuke Hayashi,  Akihiro Fujimoto,  Hideki Tode,  

pp.-

Publication Date:2020/12/2

Online ISSN:2188-5079

DOI:10.34385/proc.63.E3-1

PDF download

PayPerView

Summary:
The opportunities such that many users in local area holdthe same content to share the same experience are increasing. BitTorrent,a P2P-based file sharing software, is an effective solution for distributinglarge files to a large number of users. However, BitTorrent does not assumefile sharing between nearby users, so download efficiency could drop bycommunicating with distant users frequently. In this paper, we propose a BitTorrent-based group-aware file sharing method, in which nearby peers create a collaborative group to improvethe download efficiency. We modify two major algorithms used in BitTorrentto optimize for the collaborative groups, named “in-group rarity basedpiece selection algorithm” and “group-first choking algorithm.” Simulationresults demonstrate that the proposed method improves the downloadperformance of the collaborative group.