On Minimizing Average End-to-End Delay in P2P Live Streaming Systems

Peer-to-peer (P2P) live streaming provides a scalable solution to the distribution of multimedia content. However, existing streaming applications are plagued by the problem of long playback latency, which discourages commercial IPTV deployment from the ISP end. Moreover, ISP may provide viewers with diverse service options with different video quality, such as 720HD and 1080HD. Obtaining assurances on meeting the delay constraints in such dynamic and heterogeneous network environments is a challenge. In this paper, we devise a streaming scheme which optimizes the bandwidth allocation to achieve the minimum average end-to-end P2P streaming delay. We first develop a generic analytical framework to model the minimum average delay P2P streaming problem, called the MADPS problem. We then present iStream to solve the MADPS problem. The core part of iStream is a fast approximation algorithm, called iStream-APX, based on primal-dual schema. We prove that the performance of iStream-APX is bounded by a ratio of 1 + ω, where ω is an adjustable input parameter. Furthermore, we show that the flexibility of ω provides a trade-off between the approximation factor and the running time of iStream.