Special seminar -- Bittorrent Peer Identification Based On Behaviors of A Choke Algorithm
Detail: AJ Chalermek from Chulalongkorn University will give a seminar on Bittorrent Peer Identification Based On Behaviors of A Choke Algorithm. Please join the seminar tomorrow evening na kub at Computing Center room 401 at 6:15 pm.
Bittorrent is currently one of the most popular peer-to-peer file sharing protocols. However, it incurs such excessive amount of traffic that it may adversely affect the performance of legacy internet applications. To limit this adverse impact, an efficient methodology for bittorrent identification may be needed. This work proposes a novel approach to identify bittorrent peers based on behaviors of the choke algorithm (the main algorithm used for controlling data exchanges among bittorrent peers) instead of using the transport-layer information. Therefore, this work can identify even the peers that attempt to avoid being detected by altering their flow connection at the transport layer. Given that this work relies on only the network-layer information, our approach is capable of achieving robustness to changes in the transport layer, maintaining fewer states, and identifying peers that encrypt their traffic. The experimental results on the controlled traffic and normal traffic indicate that this approach can efficiently and quickly identify most of excessive bandwidth-consuming peers. Hence, it is possible to manage, stop, or control bittorrent peers before they can transfer a large amount of data. Furthermore, with our low false-positive rate, the approach will rarely misidentify peers and will unlikely worsen the performance of legacy internet applications if not improving it.