Research On Data Backup In Wireless Sensor Networks

With the development of embedded computing technology,wireless communication technology and sensor technology,the development of wireless sensor networks with sensing,transmitting and storage capabilities has become possible.At present,wireless sensor networks are widely used in various fields such as agriculture,industrial,medical and other fields.In practical applications,the sensor nodes exposed to the air are vulnerable to the environment and the failure of node often occurs,resulting in the incomplete data set.So the user can not effectively and accurately use the sensing data for analysis.However,WSN is data-centric,and the integrity of data is of great significance to the usability and performance of WSN.Therefore,it is very important to study the problem of perceived data backup.Analyze and summarize the existing algorithms,the data backup algorithm mainly faces the following two problems:Firstly,the sensing data collected by a large number of sensing nodes needs to be backed up,which presents a huge challenge to the energy-limited sensor nodes.Secondly,to ensure network fault tolerance,how to choose the backup node and determine the number of backups.In view of the above problems,this paper studies the data backup algorithm from the three indicators,including the average node energy consumption,data recovery rate,network lifetime.The main research work is as follows:First,we study the problem of data backup in WSN,which is based on space-time redundant data clearance is studied.Aiming at the shortcomings of the existing algorithms,this paper focuses on the set of backup data,which is studing the spatiotemporal correlation of perceptual data in wireless sensor networks,and it proposes a data backup algorithm(TS_DB)based on space-time redundant data clearance.Firstly,the algorithm divides the network into k clusters by using k-means algorithm,and then using the mining mode to explore the spatial correlation between the cluster nodes and the cluster head nodes.To determine the perceived data set to be backed up in the network,the univariate linear regression model was used to establish the time correlation of single node perception data,and within the error range,redundant data is eliminated.Finally,according to the residual energy of the node,the number of backups is determined.The experimental results show that the proposed TS DB algorithm has lower average energy consumption than the related work,and the life span of the network has been greatly improved.Second,we studied energy-efficient data backup.Based on the number of backups,this paper presents an energy-efficient data backup algorithm(EE_DB)in an energy-efficient way to ensure network fault tolerance.Considering that the sensor nodes are vulnerable to the environment,the backup node has the blindness of backup data,resulting in a large amount of redundant data in the sink node.In this paper,from the perspective of redundancy data clearance and fault tolerance,the set of robust node is introduced to back up the sensing data of regular nodes.The algorithm firstly sets up the model of space-time redundant data clearance,and then the regular nodes selects the backup path according to the weight matrix,so that the sink node select some robust nodes which can acquire all the sensing data.Experimental results show that the EE DB algorithm proposed in this paper has a greater improvement in energy and fault tolerance than the related work.