论文部分内容阅读
在高速网络控制与带宽分配研究中,网络业务尾分布分析是一个重要问题.利用大偏差理论对网络分数自回归整合滑动平均(FARIMA)业务队列尾分布进行分析,得出队列尾分布渐近服从Weibull分布.给出了一个基于Markov过程平稳分布的参数估计方法.最后对Bell实验室实测数据进行了实证分析.
In the research of high-speed network control and bandwidth allocation, the analysis of tail distribution of network services is an important issue.Using large deviation theory to analyze the tail distribution of FARIMA business queues, the tail-tail distribution is asymptotically obeyed Weibull distribution.A parametric estimation method based on the stationary distribution of Markov process is given.Finally, the experimental data of Bell laboratory are analyzed empirically.