计算机与数字工程2016，Vol.44Issue(8)：1416-1419,1517,5.DOI:10.3969/j.issn.1672-9722.2016.08.005
一种 P2P 数据流聚类的算法
Algorithm of P2P Data Flow Clustering
摘要
Abstract
P2P technology is one of the fastest network technology development today ,but a large number of P2P data flow causes some impact to internet ,so in order to control the adverse impact of P2P traffic generated by the internet ,it is necessary to identify and classify P2P and non-P2P traffic flow .However ,the data flow clustering is an important step to study dataflow of different properties ,a P2P data flow clustering algorithm is proposed based on network traffic behavior characteristics ,thinking of the algorithm is :first description of P2P traffic packets features ,according to the characteristics of P2P packets using similarity calculation algorithm similar to P2P flow calculations ,followed by the use of decision tree al-gorithm for P2P and non-P2P data flow classification ,and finally the use of Euclidean distance and PSO similar P2P flow cluster .关键词
P2P 流/粒子群/相似性/欧氏距离/决策树Key words
P2P flow/particle swarm/similarity/Euclidean distance/decision trees分类
信息技术与安全科学引用本文复制引用
朱俚治..一种 P2P 数据流聚类的算法[J].计算机与数字工程,2016,44(8):1416-1419,1517,5.基金项目
北京航空航天大学软件开发环境国家重点实验室开放基金项目（编号SKLSDE-2013KF-02）资助。 （）
