摘要
自相似性对网络性能产生了影响是当前的研究热点。建立了一种基于FBM的自相似网络排队时延抖动分析模型,重点讨论了自相似流量作为输入时对排队系统的时延抖动的影响。对理论分形流量和实际测量流量进行了仿真实验,验证了结果的正确性和有效性。实验结果表明:自相似流量长相关强弱的程度对排队系统时延抖动特性具有非常不同的影响,尤其是在缓存较大的情况下。同时,还发现网络流量中长相关发生作用时状态转变与排队系统本身的参数也有关,这是新的发现,对实时业务的网络性能评价具有重要的参考意义。
The focus of current research is how Self-similarity impacts network performance.A queuing delay jitter model, based on FBM for self-similar network, is proposed and self-similar traffic as input on the impact of the delay jitter of queuing system is discussed, which is important.Theory fractional traffic and actual measured traces are used to simulate experiment.Simulation confirms the correctness and efficiency of these results.The experimental results demonstrate that the degree of the long-range dependence for the self-similar traffic has quite a different impact on the delay jitter of queuing system, especially on the condition of a large buffer size.The state changes of long-range dependence is related to the parameters of queuing system.These findings are new, which is an important reference for the performance evaluation of real-time network .
出处
《计算机系统应用》
2010年第9期101-104,共4页
Computer Systems & Applications
基金
国家自然科学基金(60572143)
关键词
自相似流量
排队模型
时延抖动
分形布朗运动
长相关
self-similar traffic
queuing model
delay jitter
fractional brownian motion(FBM)
long-range dependence