论文部分内容阅读
提出了一种对等网络中继节点选择的随机路由算法,该算法应用随机规划框架,通过权衡本地路径真实时延以及非本地网络路径的时延统计分布,从而选择端到端期望时延最短的中继节点完成流量传输。随机中继路由算法可分布式实现,通过相邻节点动态更新路由的统计测量信息,相比于经典的静态路由算法能够获得更低的时延性能。为了更好地测量非本地网络覆盖路径的统计时延分布,路由算法拟合覆盖链路上的历史时延测量数据,并通过仿真实验表明,基于本算法建立的中继单路径/多路径可有效减少端到端路径时延和丢包率。
This paper proposes a stochastic routing algorithm for peer-to-peer network relay node selection. This algorithm uses stochastic programming framework to select the shortest end-to-end expectation delay by weighing the real time delay of local path and the delay statistical distribution of non-local network path. Of the relay node to complete the transmission of traffic. The stochastic relay routing algorithm can be implemented in distributed manner. By using the neighboring nodes to dynamically update the statistical measurement information of the routes, the stochastic relay routing algorithm can obtain lower delay performance than the classical static routing algorithm. In order to better measure the statistical delay distribution of the non-local network coverage paths, the routing algorithm fits the historical delay measurement data over the coverage links. The simulation results show that the single path / multi-path based on this algorithm can be established Effectively reduce the end-to-end path delay and packet loss rate.