论文部分内容阅读
为了缓解点对点(P2P)流媒体系统给互联网带来的通信压力,设计了一种基于P2P流媒体网络的协作缓存机制.首先,对已有的大规模P2P流媒体系统进行了测量实验,发现它们通过部署专用服务器提高系统整体性能,而节点之间的协作比较差.然后,利用缓存数据块的被访问频率计算其价值,利用数据块的传输时延计算其请求分布因子.最后,节点结合数据块价值和请求分布因子替换已缓存数据,尽可能地保留那些来自较远节点的缓存数据和高频数据,以实现缓存数据的均衡分布,提高节点之间的协作性.模拟结果显示所提出的方法在在缓存命中次数、系统负载和节点协作性等多个尺度上有较好的性能.
In order to alleviate the communication pressure brought by peer-to-peer (P2P) streaming media system to the Internet, a collaborative caching mechanism based on P2P streaming media network is designed.Firstly, the existing large-scale P2P streaming media system has been measured and found that they By deploying a dedicated server to improve the overall performance of the system, and the cooperation between the nodes is relatively poor.Then, the value of the cached data block is calculated by using the accessed frequency of the data block, and the request distribution factor is calculated by the data block transmission delay.Finally, Block value and request distribution factor to replace the cached data, as far as possible to retain those from far away from the cache data and high-frequency data in order to achieve a balanced distribution of cached data to improve the coordination between nodes.The simulation results show that the proposed The method has better performance on multiple scales such as cache hit count, system load and node cooperation.