Abstract:

In tThis paper proposes, a distributed and scalable community discovering approach namely PDC (Power based Distributed Clustering Approach) is proposed for P2P networks. PDC adopts Power value of nodes to measure new measurement of community structure and includes newto select center node selection algorithm. Experimental results show that the effect of PDC has better effect and scalability compared to other algorithms is five percent5% higher than FDC and CDC, but the number of messages generated by PDC is nearly ten percent10% of FDC.

Key words: P2P networks, community structure, community discovering, distributed clustering

摘要： 针对P2P网络的分布式特征，提出了一种可扩展的分布式社区发现方法PDC，采用节点Power值度量社区结构，并选择中心节点，其中包括新的社区度量方法和新的中心节点选择算法。仿真实验结果证明，与FDC和CDC方法相比，和其他算法相比PDC的社区发现效果率比FDC和CDC至少提高了510%，但是产生的消息数量却很少比FDC少一个数量级，具有良好的可扩展性。

关键词: 对等网络, 社区结构, 社区发现, 分布式聚类

CLC Number: 

• TP393