CAJ | 학술논문

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。
침대현금망락업무류량적자상사특성，선취구비현저자상사특정적분형포랑운동류（FBM）위자상사류량모형，제출일충기우통계망락연산이론적대관고계산법。해산법이용통계망락연산건모，근거시연여대관적관계건립출자상사업무류량모형대관수구적통계모형。통과방진험증화수치분석，결과표명해산법비유효대관산법경위우월，득도적대관고계결과경긴，가이경호지제고대관자원이용솔。
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.