Research On Mesh-based P2P Live Streaming Technology In Heterogeneous Environment

With the increasing popularity of the Internet and the rapid development of multimedia technology, video-based streaming services will become the Internet's killer application. However, it's a great challenge to solve the problems of high bandwidth demand caused by the huge video data, long-lasting service and strict real-time demand.As the advantages of superior scalability and low cost, ease of deployment, Peer-to-Peer becomes one of the most important technology solutions to large-scale streaming media applications. After doing detail research on most achievements of recent P2P streaming media technology, this paper determines the two most key problems on this field: the structure of overlay network and data segment scheduling policy. The former mainly concerns about the problem of how allocate neighbors to peer, the latter solves the problem of how efficiently transfer data segments among peers in the neighbor. This paper focus on them to expand research.Firstly, by researching the literatures with regarding to data segments scheduling algorithms, it is founded that they consider the heterogeneity of practical network condition seldom, viewing all peers the service ability as the same. In order to make full use of the heterogeneity to improve the performance of system. This paper introduced the idea of bandwidth-aware. At the mean time, this paper designed three kinds of bandwidth-aware data scheduling algorithm which represent different node upload bandwidth dependency to study bandwidth-aware based scheduling algorithm deeply. Simulation results show that the performance of system will improve to some extent if the characteristics of node heterogeneity can be exploited reasonable——preferring high upload bandwidth node to receive data.Meanwhile, the majority literatures consider overlay network as static random or fully connectivity status. Also parts of them pose the concept of preference, selecting low transport delay and high upload bandwidth peer as neighbors, it improves system performance to some extent compared to random topology. This paper proposed an overlay topology adaptive optimization algorithm, which is simple, distributed and with the ability of measuring node upload bandwidth directly. Simulation results show that it can improve the performance of system efficiently and has certain robustness.