Self-Adaptive Distributed Data Storage Scheme Based On Projection Regeneration Coding

Taking advantages of the diversity,adaptability and reliability of sensors in the network,and the excellent features such as rapid deployment,self-organization,high fault tolerance,and high concealment of the network,the Wireless Sensor Networks(WSN)are widely used in many fields.However,due to the large number of sensor nodes in the network and the large network size,the WSN is still faced with many problems like quickly network topology changes,complex network deployment procedures,weak node data processing capabilities,and limited node energy.Appling the Compressed Sensing(CS)theory to the distributed WSN to achieve efficient compression of sensor node's data.CS can reduce the amount of data processing under the premise of better reconstruction signal quality,and then reduce the hardware requirements for transmission.However,CS's high signal compression ratio still cannot solve the problem of blind spots and caverns,which caused by dynamic changes in the network topology and high packet loss rate.Regenerating Coding(RC)is an extension of distributed network coding(NC)technology.The high network throughput,low transmission delay,and high bandwidth utilization of RC can exactly alleviate the adverse effects caused by dynamic network topology changes in the WSN.Therefore,how to combine CS with RC and design an efficient data fusion scheme for sensor nodes to reduce the data traffic and energy consumption of distributed nodes in the network is a problem in the WSN.This paper firstly combines CS information acquisition theory with RC information interact technology,proposes a new Projection Regeneration Coding(PRC)method,gives the physical definition and basic principle,and expounds its projection characteristics and regeneration characteristics.As a kind of information acquisition and exchange technology that combines sampling,routing and coding,PRC not only has the sparseness and compressibility of CS,but also can accurately reconstruct the best approximation of the signal through a few sub-projection measurements.It also features RC's high throughput,high bandwidth utilization,and balanced network load.While solving the problems of link instability and loss-prone packets in WSN,PRC can also reduce the nodes' computational complexity,transmission energy consumption and extend the nodes' service life.Secondly,using the clustering theory of WSN and the spatial-temporal correlation of sensornode readings,proposing an inter-cluster neighborhood assisted projection regeneration coding(C-NPRC)On the one hand,this scheme combines PRC basic theory to achieve projection compression and regeneration coding of sensor node data in clusters.On the other hand,it uses neighbor assistance to achieve the fusion of cluster head node's data between clusters and clusters.As the applied research on the network level of distributed WSN,under the premise of guaranteeing the reconstruction accuracy of the original signal,this scheme can effectively improve the signal's compression ratio and reduce the transmission cost and energy consumptionFinally,combining the basic theory of PRC and the codec concept of C-NPRC,this paper studies the adaptive distributed data storage methods,which are applicable to the dynamic changes of WSN sensor nodes readings,and proposes a self-adaptive distributed data storage scheme based on projection regeneration coding(ST-PRC).In the ST-PRC method,the relay nodes in the cluster can adaptively adjust the forwarding probability to meet the dynamic change requirements of the convergence nodes for collecting data accuracy.The solution can meet the dynamic network adaptive storage-access requirements and effectively balance the internal resources and energy consumption of the network.In the premise that the convergence precision of the access data of the sink node can meet the needs of the user,it can effectively reduce the cost of data exchange,thereby reducing the overhead in the network and saving network energy consumption.