Adaptive Replication Management Algorithm Based on Location and File Popularity for Peer-to-Peer Network

CHEN Kang,YU Hong-Liang,ZHANG Kun
DOI: https://doi.org/10.3724/SP.J.1016.2009.01927
2009-01-01
Chinese Journal of Computers
Abstract:File sharing is one of the most important applications based on the peer-to-peer network structure. In such application,the data transferring speed is more important than the network latency. The authors have done some work on the replication algorithms based on the peer-to-peer network. Such algorithms are meaningful for improving the reliability of network applications as well as reducing the network traffic. For improving the performance of file sharing with reduced network traffic,this paper proposes a method called PLAR (Popularity and Locality-based Adaptive Replication). PLAR is based on the locality and popularity of objects as well as applying the mixed server selection method and remote boosting schemas. PLAR is used in the authors’ Granary system which is data storage system for wide area network. The experiment results show that PLAR can improve the download speed over 60% on average which means that it can improve the sharing as well as reducing the network traffic.
What problem does this paper attempt to address?