基于复杂网络拓扑结构的评估体系研究  

Research on the assessment system based on complex network topological structures

作　　者：邵铁林 

机构地区：[1]沈阳理工大学信息科学与工程学院,辽宁沈阳110159

出　　处：《微型机与应用》2017年第1期69-72,共4页Microcomputer & Its Applications

摘　　要：对复杂网络拓扑结构中特定属性的计算和评估进行了研究。层次分析法和熵权法被用来计算复杂网络拓扑结构的权重,这样既可以防止在主观赋权法上因专家经验不足使得对方案的排序造成很大的随意性,又可以避免客观赋权法的结果与实际情况不一致的现象发生。选取了规则网络模型、随机网络模型、小世界网络模型和BA无标度网络模型进行加权计算,并计算出它们各自相应的综合权重。从综合权重的角度来看,实验结果表明,随机网络效能最好,接下来是小世界网络、无标度网络,最后是规则网络。该评估体系和算法可以被广泛地用于需要对网络性能进行定量评估的情况下。A quantitative way of measuring and evaluating specific properties in complex networks topology is studied. By determining the weight of the complex network topology structure, both hierarchy process and entropy weight method were introduced to avoid incorrect results out of great randomness caused by experts who are lacking of experiences in the subjective weighting method, as well as to avoid the inconsis- tent output of objective weighting method with that of the real situations. Weighting method was performed on four different complex networks models including regular network model, random network model, small world network model and BA scale-free network model, and comprehen- sive weights of corresponding networks were ealculated. The experimental results show that on comprehensive performance point of view, ran- dom network is the best followed by small world network, scale free network rank next and the regular network at last. The assessment systems and algorithms could be widely used in circumstances where quantitative evaluations of networks performances are required.

关 键 词：主观赋权法 客观赋权法 组合赋权法 网络性能评价 

分 类 号：TP302.7[自动化与计算机技术—计算机系统结构]