无线传感器网络含有缺失数据的数据融合研究

Data fusion research based missing data in wireless sensor networks

针对含有缺失数据的无线传感器网络数据融合问题,提出了一种简单易行的二次数据融合算法(TFA)。考虑到感知数据的时—空相关性,对缺失数据进行线性插值插补和回归分析插补,对插补结果利用线性组合融合算法进行融合。综合考虑各节点的信息,利用自适应加权融合算法进行融合。实验结果表明,该算法在含有缺失数据的前提下,能够以较低的计算开销和较高的估计精度实现数据融合。

This paper proposed a simple and feasible twice data fusion algorithm（TFA） to solve the problem of data fusion based data missing.As considered the sensing data＇s temporal and spatial correlation,it used the linear interpolation imputation algorithm and the regression analysis imputation algorithm to estimate the missing data,then used an adaptive linear combination fusion algorithm to estimate.As considered the data of each node,it also used an adaptive weighted fusion algorithm to estimate.The simulation results show that the algorithm has lower computing cost and better fusion effect with data missing case.