基于TEEN协议和BP神经网络的WSN数据融合模型

摘要

针对无线传感网感知数据中含有大量无效或冗余数据的现象,提出了一种基于TEEN协议和BP(back propagation)神经网络的数据融合模型.该模型利用三层BP神经网络描述簇结构,通过TEEN阈值过滤非必要信息,在簇结构信息传输过程中运用神经网络功能函数处理大量感知数据,从中提取感知数据的特征值并转发至汇聚节点.实验仿真表明,该模型无论在数据通信量、使用寿命及网络消耗上都优于TEEN协议,在降低网络通信量和网络能耗的同时提升了网络的使用寿命,大大提升了数据采集的效率和性能.%In view of the phenomenon that the wireless sensor network has a large number of invalid or redundant data, this paper proposed a data fusion model based on TEEN protocol and BP (back propagation) neural network.The model used three layer-based BP neural network to describe the cluster structure, and filtered unnecessary information through a TEEN threshold.During the process of information transmission, the performance function of neural network was used to deal with large amounts of sensing data, where feature value of sensing data was extracted and transmitted to the sink node.Experimental results show that the proposed model is superior to the TEEN protocol on data traffic, life cycle and network consumption.As a result, the life cycle of the proposed model is improved while reducing network traffic and network energy consumption.Hence, both the efficiency and performance of data collection are greatly improved.