摘要： 针对对等(P2P)流媒体直播系统中节点的异构性问题，将系统节点分为不同级别，基于随机行走算法选择邻居节点，采用Metropolis- Hastings算法计算随机行走的转移矩阵，从而保证节点能以较大的概率选择到系统中级别较高的节点，并进一步描述覆盖网拓扑优化算法。仿真实验结果表明，该算法能促进节点协作，降低系统延时，在动态网络环境下具有良好的鲁棒性。

关键词: 对等网络, 流媒体直播系统, 随机行走, 拓扑优化, 邻居选择

Abstract: Considering the heterogeneous features of the peer in Peer-to-Peer(P2P) media streaming system. In this paper, the system nodes are divided into different levels. A neighbor peer selection algorithm is proposed based on random walk, and the transition probability matrix is calculated through the Metropolis-Hastings methods in order to choose a higher level of system nodes with greater probability. Moreover, the overlay network topology optimization algorithm is elaborated. Simulation results demonstrate that the algorithm can improve the peers’ cooperation, reduce system latency, and has a strong robustness in dynamic network environment.

Key words: Peer-to-Peer(P2P) network, media streaming live system, random walk, topology optimization, neighbor selection

中图分类号: 

• TP393