摘要
基于线形分形稳定噪声(LFSN)的网络业务流模型可以很好地描述流量的自相似和重尾特性,对这种模型的随机突发边界进行理论推导,得出较现有文献更为普遍的结论,尤其是其随机上界更为精确,因此在随机网络演算分析中有着重要意义。同时基于对一般LFSN过程的快速模拟,设计了一种独特的实验方法,得到突发边界随机分布的估计值,验证了理论推导的正确性。
A stochastic burstiness boundary of general LFSN（linear fractional stable noise）-based traffic model,which is an important model that has been proven capable to capture both properties of self-similar and heavy tailed,was derived.The final result is much more general than the current one,especially for the upper boundary which is also more accurate.Meanwhile,a unique experiment based on fast simulation of general LFSN process was set up to get an estimation result for actual distribution,which in turn proved theoretical derivation correct.
出处
《通信学报》
EI
CSCD
北大核心
2010年第5期16-21,共6页
Journal on Communications
基金
国家自然科学基金资助项目(60972016
60872010)
新世纪优秀人才支持计划(070339)
国家高技术研究发展计划("863"计划)基金资助项目(2009AA01Z205)
国家实验室基金资助项目(P080010)~~
关键词
随机突发边界
自相似
随机网络演算
性能分析
线性分形稳定噪声
stochastic burstiness boundary
self-similar
stochastic network calculus
performance analysis
linear fractional stable noise
