Research On Data Quality Control Of Wireless Sensor Networks

Randomly distributed sensor nodes form a network in self-organized way by means of low-power wireless communication sense the environment information and send the data to the base station through multi-hop forwarding.Generally,these nodes are referred to as wireless sensor networks(Wireless Sensor network,WSN).Wireless sensor networks need to constantly collect data from the physical world and transfer,process,analyze and share the data.Wireless sensor network to provide effective service is on the premise of ensuring the quality of data nodes,so the node data quality of perceived data is a concern.This paper focuses on the two most important dimensions of WSN data quality:the correctness dimension and the integrity dimension.In order to improve the correctness of WSN data quality,the data denoising method is used to reduce the influence of invalid observation data and excessive deviation observation data on sensor network data quality to a certain extent.On the other hand,in order to guarantee the integrity of WSN data quality,we propose an effective loss data estimation algorithm to repair the lost data,and make the data integrity of the sensor network to a large extent under the premise of guaranteeing the accuracy of the repair data.In this paper,considering the low cost of the hardware and the complexity of the deployment environment,it is possible to delay or lose some data when transferring data between nodes.This paper mainly studies how to estimate the lost data of the node to ensure the quality of the data,and then guarantee the quality of the network.Data loss estimation algorithm based on gray correlation degree and Data loss estimation algorithm based on spatio-temporal correlation degree.Data loss estimation algorithm based on gray correlation degree:the algorithm is mainly divided into data denoising,gray correlation degree calculation,neighbor node selection,estimation of missing data.The factors of the monitoring environment,resulting in noise has affected the node sensing data.When selecting the neighbor nodes of missing data nodes,the sensing data of the neighbor nodes need to be denoised.In this paper,we select the median filter of image denoising algorithm.The main idea is regarding the sensor nodes as the pixels in the image,regarding the node sensing data as the pixel value.Then calculating the gray correlation degree between the node and its neighbor nodes,and select the neighbor nodes with larger correlation degree.Finally,the data of the adjacent nodes are used to estimate the lost data according to the high correlation degree.Experimental results show that the algorithm proposed in this paper has higher accuracy than the related algorithms.Data loss estimation algorithm based on spatio-temporal correlation:The algorithm firstly performs wavelet denoising on node-aware data,and then estimate the data in time correlation degree using the Markov chain.The spatial correlation algorithm is used to estimate data when the time correlation algorithm is invalid.In the process of spatial correlation,the length of the data sequence is not consistent caused by the data loss of two nodes.The DTW algorithm is used to calculate the correlation degree of the two nodes.The spatial correlation process is divided into the similarity distance calculation,the DTW correlation matching,neighbor node selection and data estimation.The experimental results show that the proposed algorithm has higher accuracy and robustness than the related algorithms.