基于图谱超分辨的WSN节点间数据填充方法

Data filling method between WSN nodes based on super-resolution map spectrum

为更好地解决无线传感器网络(WSN)节点间的数据缺失问题,快速准确地提高节点探测的精度与质量,提出一种基于图谱超分辨的WSN节点间数据填充方法。从数据处理角度出发,将WSN节点间数据填充问题转化为数据的图谱超分辨问题来处理,即将WSN节点数据进行二维图谱化,考虑传感器测量数据的特点,建立合适的归一化图谱模型;利用马尔可夫随机场模型来处理建立的无线传感器图谱,实现对节点间空白区域数据的准确估计。实验结果表明,与3种经典的WSN缺失值估算方法相比,所提方法可获得较高峰值信噪比(PSNR)和较低根均方误差(RMSE)。

To effectively solve the problem of data loss between nodes in wireless sensor networks(WSN)and quickly and accurately improve the accuracy and quality of the node detection,a data filling method between WSN nodes based on the super-resolution map spectrum was proposed.From the data processing point perspective,the WSN nodes data were transformed into super-resolution map data,achieving two-dimensional map data of WSN nodes.The characteristics of sensor data were considered and an appropriate normalization map spectrum model was established.Markov random field model was used to deal with the established WSN map spectrum data,which achieved an accurate estimation of the blank area between data nodes.Simulation results show that compared with three classical WSN missing value estimation methods,the proposed method has higher peak signal to noise ratio(PSNR)and lower root mean square error(RMSE).