Foresighted Resource Reciprocation Strategy in BitTorrent Systems

Recent research efforts have shown that the popular Bit-Torrent protocol does not provide fair resource reciprocation and allows free-riding. In this paper, we propose a novel foresighted resource reciprocation mechanism that replaces the peer selection mechanism with a reinforcement learning mechanism that adopts a foresighted resource reciprocation policy. We model the peer interactions in the BitTorrent-like network as a stochastic-game, where we explicitly consider the strategic behavior peers. The peers can observe partial historic information of associated peers’ statistical reciprocal behaviors, through which the peers can estimate the impact on their expected utility and then adopt their best response. The policy determines the peer’s optimal resource reciprocations, and enables the peer to maximize the long-term performance. The mechanism improves fairness as it relies on long-term history. Moreover, it hurts the free-riders directly since foresighted peers are discouraged to upload to free-riders. We have implemented the proposed mechanism in an existing BitTorrent client. We have also performed extensive experiments on a controlled PlanetLab testbed to evaluate the mechanism effectiveness. Our results confirm that the foresighted resource reciprocation mechanism promotes fairness, improves the system robustness, and discourages free-riding in compare to the regular BitTorrent.