Data fusion algorithm analysis and realization based on wireless sensor networks
COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(12B) 363-368
CAPF Of Engineering University, wujing road NO 1,Weiyang District, Xi'an City, Shaanxi Province, 710086, China
In the premise of ensuring the veracity of model prediction result as well as simplify the model and prediction algorithm, this paper introduces a wireless sensor network data fusion algorithm based on ARMA time series model. This algorithm aims at reducing the energy consumption of wireless sensor network and improving the accuracy of fusion result. It conducts reliability analysis to node collecting data and removes the abnormal data. By analyzing ARMA model, we find that the construction of prediction model and cost of predicting is related to order of the model. The experiment result shows that this algorithm can not only reduce the network energy consumption but also detect abnormal data. ARMA model that determinates by BIC&F applicability test methods can adapt the wireless sensor network well.