【摘要】 虽然现有的信誉机制减少了P2P文件共享系统中的"搭便车"和"公有悲剧"现象,但依据信誉大小来选择服务节点仍会造成网络拥塞以及"马太效应"等问题.为此提出了一种基于用户下载需求偏好的P2P信誉模型,该模型将服务节点中所请求文件的评价值和其承诺提供的带宽大小这两种因素相结合,作为请求节点选择服务节点的依据.根据不同用户偏好的效用大小不同,服务节点可以动态制定当前的承诺带宽以吸引其他节点下载.仿真结果表明,该方案可以有效缓解热门节点造成的网络拥塞现象,并可以提高网络的整体资源利用率.更多还原

【Abstract】 Though traditional reputation mechanisms can reduce the phenomena of "free-riding" and "tragedy of the commons" in P2P file-sharing systems,selecting server peers according to reputation suffers from network congestion and the "Matthew effect".A preference for downloading requirements-based P2P reputation model was proposed.This model combined two factors as the basis for selecting server peers:the evaluation of the requested file and bandwidth commitment.In view of the fact that different users preferences lead to different utilities,a server peer could allocate the bandwidth commitment dynamically to attract request peers.The simulation shows that the method can ease the congestion and raise the overall rate of utilization of resources in P2P file-sharing systems.更多还原