Optimal TCP-Friendly Rate Control for P2P Streaming: An Economic Approach

TCP and TCP-friendly rate control protocols, designed for unicast, do not take neighbor connections into account in P2P networks. In this paper, we study the topic of distributed and optimal rate control for scalable video streams in P2P streaming applications. First, we propose a fully distributed and TCP-friendly network analytical model for rate control and formulate an optimization problem to maximize the aggregate utility for the P2P streams. In the model, we further extend the definition of TCP-friendliness for P2P network. Second, we propose a shadow price-based distributed algorithm for P2P Streaming that solves the optimization problem. Finally, we evaluate the performance of the proposed algorithm in terms of streaming quality and messaging overhead. Extensive simulations show that the proposed algorithms generate very small overhead and that they are optimal in terms of overall quality for scalable streams.