机会网络动态社团的预测

由于人们之间社会关系相对稳定并且存在一定的依赖性,由人携带设备组成的机会网络中会出现节点的聚集现象,从而表现出很好的社团特性.提出一种应用贝叶斯-蒙特卡洛(Bayesian-MCMC)预测机会网络节点社团分配的新方法,并在两个不同地点的机会网络数据集上对该方法进行了评估,实验结果显示,此方法能对机会网络中的社团演变进行预测,达到了很高的准确率,且具有良好的鲁棒性.对机会网络社团快速准确的预测有利于机会网络中节点的管理,消息的传输,资源的分配,并可以为探索由人携带设备组成的机会网络这类场景的移动模型的数学分析提供理论依据.

Forecasting Dynamic Communities in Opportunistic Networks

As a result of the relative stability and the dependence of social relations between people,the nodes in opportunistic networks appear clustering phenomena and the opportunistic networks demonstrated good community properties.A new method for predicting the distribution of node′s community,based on Bayesian-Monte Carlo(Bayesian-MCMC),is proposed in this paper.This method is evaluated through opportunistic network′s data collected from two different locations.Experimental results show that this method can detect communities in the opportunistic networks and predict its time evolution,which achieve a high accuracy and high robustness.Quickly and accurately detecting and forecasting the communities in opportunistic networks are beneficial to network node management,message transmission,resources allocation and providing a theoretical background of exploring the mathematical model on opportunistic networks constituted by people carrying equipment.