中文摘要：

通过对网络模拟流量模型的分析,深入研究了基于时间驱动流量模型的流模拟方式,并针对目前基于时间驱动的流量模型time step选取方法存在的缺陷,提出一种根据网络行为的复杂程度不断修改time step的自适应time step选取策略.实验数据表明,相比于目前流量模型time step选取策略,这种新的选取策略大幅度降低了网络模拟的执行时间,同时提高了模拟的精确度,对大规模的网络模拟也具有很好的效果,为建设大规模高真实性的实时网络模拟平台奠定了基础.

英文摘要：

The network fluid simulation of time-driven fluid models was deeply studied through the analysis of present fluid models for network simulation,and considering the deficiencies of current time-driven fluid models in time step selection,a new algorithm for adaptive selection of time step was proposed which adjusts the time step continuously according to the complexity of network behaviors.The experimental results show that the adaptive time step selection algorithm can reduce the execution time for network simulation obviously while increasing the simulation accuracy compared with the other strategies.Furthermore,good effects can also be obtained for large-scale networks when using the adaptive time step selection algorithm,which will provide the foundation for studying the large-scale realtime network simulations.