一种新的物联网传感器数据融合方法

A new data fusion method for IoT sensors

为了更好地对物联网中的无线传感器网络进行节能处理,需要对网络中的设备进行集群操作。文中提出了一种新的聚类方法,该方法通过多种传感器模式开发了一种节能的数据聚合协议。它使用融合相似矩阵将网络中的设备分成不同的簇,使用建模误差在置信区间内计算不同模式融合的权重。提出的聚类方法是使用分层凝聚聚类框架。实验结果表明,文中提出的方法具有较好的效果。

In order to better perform energy-saving processing for wireless sensor networks in the Internet of Things,it is necessary to perform cluster operations on devices in the network.This paper presents a new clustering method,which develops an energy-saving data aggregation protocol through multiple sensor modes.It uses a fusion similarity matrix to divide the devices in the network into different clusters.Then,the modeling error is used to calculate the weight of the fusion of different models within the confidence interval.The proposed clustering method uses a hierarchical agglomerative clustering framework.The experiment results show that the method proposed in this paper has better results.