Title :
Analysis of mobile BitTorrent client behavior
Author :
Csorba, Kristof ; Ekler, Peter ; Bori, Andras ; Charaf, Hassan
Author_Institution :
Dept. of Autom. & Appl. Inf., Budapest Univ. of Technol. & Econ., Budapest, Hungary
Abstract :
In this paper we analyze the anonymous reports provided by our Android based BitTorrent client called DrTorrent. This application is available for free and is a full featured torrent client. The reports allow us to analyze the usage patterns typical for mobile torrent clients. As the mobile devices have some special properties like processing power, energy, and bandwidth limitations, the behavior differs from the traditional desktop clients. Beside the several statistics on for instance the sizes of accessed torrents and download times, we present parametric estimation models for the number of peers for a given torrent, and for the rate of peers which can be successfully connected to. This research can be considered as a pioneer work in the field of analyzing large peer-to-peer networks in modern mobile environments. The described measurement results enable to understand the mobile BitTorrent behavior in details and the proposed models can be used also as a base model for designing other mobile peer-to-peer services.
Keywords :
Android (operating system); mobile computing; parameter estimation; peer-to-peer computing; statistical analysis; Android based BitTorrent client; DrTorrent; large peer-to-peer networks; mobile BitTorrent client behavior; mobile devices; mobile peer-to-peer services; parametric estimation models; Approximation methods; Histograms; Mobile communication; Peer-to-peer computing; Protocols; Smart phones; Android; BitTorrent; parametric approximation; smartphone; usage statistics;
Conference_Titel :
Cognitive Infocommunications (CogInfoCom), 2013 IEEE 4th International Conference on
Conference_Location :
Budapest
Print_ISBN :
978-1-4799-1543-9
DOI :
10.1109/CogInfoCom.2013.6719175