基于深度学习模型的无线传感器网络数据融合算法  

Data Aggregation in Wireless Sensor Network Based on Deep Learning Model

作　　者：邱立达[1] 刘天键[1] 林南[1] 黄章超[2] 

机构地区：[1]闽江学院物理学与电子信息工程系,福州350108 [2]厦门理工学院光电与通信工程学院,福建厦门361024

出　　处：《传感技术学报》2014年第12期1704-1709,共6页Chinese Journal of Sensors and Actuators

基　　金：福建省教育厅科技项目(JA12263);福州市科技计划项目(2013-G-86)

摘　　要：为了在无线传感器网络中提高数据融合性能,基于深度学习模型,提出一种将层叠自动编码器(SAE)和分簇协议相结合的数据融合算法SAEMDA,该算法在各个簇内构建特征提取分类模型SAEM,通过SAEM对节点数据进行特征提取和分类,之后将同类特征融合并发送给汇聚节点。SAEM的训练既可以采用离线有监督学习也可以采用在线无监督学习。仿真实验表明:和BPFDA,SOFMDA算法相比,SAEMDA在网络能耗大致相当的情况下能将数据融合正确率提高最多7.5%。In order to improve the performance of data fusion in wireless sensor network,a data aggregation algorithm SAEMDA（ stacked autoencoder model data fusion algorithm ） based on deep learning model was proposed, which combined stacked autoencoder（ SAE） and wireless sensor network clustering routing protocol. Feature extraction and classification model（ SAEM） is designed by SAEMDA to extract and classify the data features of nodes in each clus-ter,and then SAEMDA sends the features fused in the same class to Sink node. Either offline supervised learning al-gorithm or online unsupervised learning algorithm can be used to train the SAEM. Simulation results show that com-pared with BPFDA and SOFMDA,SAEMDA can improve the data fusion accuracy by 7.5 percentage points at most in similar situations of energy consumption.

关 键 词：无线传感器网络 数据融合 深度学习 自动编码器 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]