移动分组网络终端数据自适应整合方法仿真

摘要

A self-adaptive integration method for terminal data in mobile packet network is proposed.Firstly,the data density relevance method is used to label the type of tag data object of each terminal neighbor node,and then the initial clustering center of terminal data category is given.Moreover,the aggregation degree in different types of terminal data is obtained,and the distance between clustering centers is got.The maximum density is taken as the first clustering center,which is put into the clustering center.The data in each terminal sensor data are estimated in batches to get the integrated optimal estimate value of any type of terminal data over a period of time.According to the optimal allocation principle of weights,the self-adaptive weighted integration within group is performed on data in each terminal sensor.From simulation results,we can see that the accuracy of data integration of proposed method is very high,which lays a foundation for improving the service quality of mobile packet network.%为了更好地提升移动分组网络的服务质量,需要进行终端数据自适应整合方法的研究.但是采用当前方法进行终端数据整合时,无法对各类数据进行聚集,存在数据整合可靠性低的问题.为解决上述问题,提出一种基于分批估计的移动分组网络终端数据自适应整合方法.上述方法先利用数据密度相关度方法标记各个终端邻居节点数据对象的类型,给出终端数据类别的初始聚类中心,获取不同类型终端数据的聚集程度,得到聚类中心之间的距离,将密度最大值作为第1个聚类中心放入聚类中心数组,对各个终端传感器的数据进行分批估计得出任意类型终端数据在一段时间内的整合最优估计值,依据权值最优分配原则对每组终端传感器数据进行组内自适应加权整合.仿真证明,所提方法数据整合精度高,为提升移动分组网络服务质量奠定了基础.