Credit Reputation Propagation: A Strategy to Curb Free-Riding in a Large BitTorrent Swarm

BitTorrent ensures cooperation among peers through its inbuilt collaborative mechanisms, however due to lack of proper incentives, significant amount of free riding is observed in it. Though in the existing literature, there exists some strategies to prevent free-riding, however, in case of large swarm sizes, it can be shown that they fail to stop free riding attempts effectively. To overcome this limitation, this paper presents a novel approach based on propagating the knowledge about the existence of possible free riders in the form of reputation among the peers within the swarm. It is shown, how a possible free riding attempt on a peer is reported to others and how this knowledge is utilized in deciding whether to upload to a particular peer or not within a BitTorrent swarm. Simulation results demonstrate how the proposed strategy effectively punishes free riders even in large swarm sizes.