节点归属性动态估计的机会网络社区检测策略

Node-belongingness dynamic estimate community detect strategy in opportunistic networks

为了提高机会网络社区结构检测的合理性和有效性,根据机会网络的特性,提出了一种低开销、分布式动态社区检测策略。根据节点的社会属性,节点动态地估计自身归属性,进而感知对所属社区的归属性,约束标签的传播过程,最终完成机会网络的社区结构检测。并将仿真结果与其他算法进行比较,本机制社区检测准确率相较于HCDA提高大约50%,且具有较强的扩展性,适用于各种复杂的网络场景。

In order to improve community structure detection of rationality and validity in opportunistic networks. According to the characteristics of the opportunistic networks. A low overhead and distributed dynamic community detection strategy is presented. According to the social attribute of nodes, their belongingness is estimated dynamically. Further, the perception of the communities belongingness can restrain the spread of labeling process. Finally, the community structure in opportunistic network is detected. The simulation result is compared with other algorithms, the accuracy of detection algorithm proposed is higher than HCDA about 50%, and it has strong scalability. The algorithm is suitable for a variety of complex network scenarios.