一种自相似流量模型的带宽估计新算法  

A NEW BANDWIDTH ESTIMATION ALGORITHM FOR SELF-SIMILAR TRAFFIC MODEL

作　　者：王程锦[1] 张中荃[1] 韩悦[1] 谢国益[1] 

机构地区：[1]解放军西安通信学院信息服务系,陕西西安710106

出　　处：《计算机应用与软件》2014年第3期112-115,共4页Computer Applications and Software

基　　金：陕西省自然科学基金项目(2011JM8033)

摘　　要：针对现今网络业务流量的自相似特性,选取具备显著自相似特征的分形布朗运动流(FBM)为自相似流量模型,提出一种基于统计网络演算理论的带宽估计算法。该算法利用统计网络演算建模,根据时延与带宽的关系建立出自相似业务流量模型带宽需求的统计模型。通过仿真验证和数值分析,结果表明该算法比有效带宽算法更为优越,得到的带宽估计结果更紧,可以更好地提高带宽资源利用率。According to the self-similarity of current network traffic, we select the fractional Brownian motion （fBm） flow, which has remarkable self-similarity feature, as the self-similar traffic model and propose a bandwidth estimation algorithm based on statistical network calculus theory. This algorithm adopts statistical network calculus theory to build the statistical model of self-similar business bandwidth demand according to the relationship between delay and bandwidth. Through simulation verification and numerical analysis, the results show that this algorithm is superior to the effective bandwidth algorithm in closer bandwidth estimation result, and can better improve the utilisation of bandwidth resource.

关 键 词：统计网络演算 自相似 分形布朗运动 带宽估计 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]