摘要
P2P网络中的节点重要性评价在实际应用中有重要意义。现有的一些重要性评价指标如度、紧密度等存在度量结果较为片面等缺点,单一指标并不能有效地对P2P网络中的节点重要性进行度量。针对此问题,提出了一种基于多重属性的综合度量指标。以节点度作为对比基础,计算多个度量指标与节点度指标的肯德尔相关系数,通过分析多个指标之间的内在关联性,最终选择节点度、介数、K-核和Page Rank四个属性来进行节点重要性的综合度量。通过真实的P2P网络进行实验验证,结果表明,相对原有单一度量指标和现有的综合度量指标,该综合指标更能够有效地对P2P网络中的节点重要性进行度量,而且具有较强的普适性,可以应用到大多数P2P网络中。
P2P network node importance measurement is of importance in practical application. Due to that the evaluation results based on some existing node importance evaluation metrics have the shortcoming of one-sidedness, therefore the importance of nodes in P2 P network can＇t be evaluated effectively by single metric. To solve this problem, a comprehensive metric based on multi-attribute was proposed. Node degree was selected as the basis of the comparision, the inherent relevance between various metrics was firstly analyzed through the Kendall correlation coefficient compution and a comprehensive evaluation metric based on degree, betweenness, K-core and Page Rank were used to compute the comprehensive metric.Through the experiment verification by real P2 P network, the comprehensive evaluation metric can more effectively evaluate the node importance in P2 P network when compared to original single metrics and existing comprehensive metrics. Also the comprehensive metrics has so strong universality that can be used to most other P2 P networks.
出处
《计算机应用》
CSCD
北大核心
2014年第A02期7-10,19,共5页
journal of Computer Applications
基金
国家自然科学基金资助项目(61170286
61202486)
