Socially-optimal ISP-aware P2P Content Distribution via a Primal-Dual Approach

Abstract

Peer-to-peer (P2P) technology is popularly exploited to enable large-scale content distribution (e.g., live and on-demand video streaming) with low server costs. Most P2P protocols in place are network agnostic, where peers download content chunks from each other regardless of their ISP belonging, leading to significant amounts of inter-ISP traffic. It has been a daunting challenge how to design protocols that optimize the P2P content distribution topology, such that inter-ISP traffic is minimized while the dissemination performance is maximized, not to mention one that motivates peer's voluntary ISP-aware peer selection. In this paper, we formulate a social welfare maximization framework for dynamical construction of the P2P content distribution topology % (namely deciding who is downloading which chunk from whom), taking into consideration both peers' gain due to chunk downloading and inter-ISP traffic that is incurred. Given its nature of an assignment problem, we resort to a primal-dual framework to design the solution algorithm, which can be practically implemented as a set of distributed, interleaving auctions, where peers bid for bandwidth to download chunks at other peers considering the potential inter-ISP traffic as a cost factor. Such an auction-based mechanism encourages peers to download from neighbors with low network costs in between, in order to succeed in bandwidth acquisition for chunk downloading. We analyze and prove the social optimality achieved by the distributed auctions and verify the performance of our proposal using realistic emulation experiments with real P2P traffic.