Toward Awareness Locality Algorithms of Peer-to-Peer File Sharing Network

In this work, we attempt to investigate the impact of locality awareness algorithm implemented in P2P file sharing application. BitTorrent has dominated in the field of P2P file sharing application. It selects node randomly from content swarm. Developers constructed a new version of BitTorrent that implemented locality awareness. This version is called TopBT. We will investigate the locality in TopBT and we will measure the QoS of TopBT and compare it to the QoS of the pure and original BitTorrent client. To facilitate our work, we constructed an AS hops count extractor tool to extract the AS hops count path between any two IP addresses. We utilized a network packet sniffer to harvest packets from TopBT and BitTorrent swarms and we used the collected data as input to our tool to extract average AS hops count paths in these swarms. This method is used to investigate the implemented locality awareness algorithm and its efficiency. Our results demonstrated that TopBT locality method reduced the average AS hops count paths. In addition, we noticed that QoS may be affected by implementing locality awareness especially for unpopular files.