摘要
通过对CAIDA多监测点获得的测量数据的分析,提取出表征Internet路由级拓扑规律的多项特征值.在特征值分析的基础上,综合考虑了路由级拓扑的动态变化,拓扑的层次性,以及节点间非线性连接的特点,提出了面向Internet路由级拓扑的DNLH(Dynamic Non-Linear Hierarchy)模型,并给出建模算法.以实际数据为依据,结合遗传算法优化了节点间非线性连接参数.通过实验分析,证明了该建模方法的有效性.
The measuring data which were got by CAIDA multi monitors were analyzed, then multi eigenvalues describing the router-level topology of Internet were extracted from analysis results. Considering the dynamic change, the hierarchy and the non-linear connecting character between the nodes in the router-level topology of Internet, the DNLH(dynamic non-linear hierarchy) model and the algorithm of the model were put forward. According to factual data, the non-linear connecting parameters between the nodes were opthnized by genetic algorithm. Through quantitative experimenting analysis, this model is found valuable.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2008年第1期57-63,共7页
Acta Electronica Sinica
基金
国家自然科学基金(No.69873007)