Research On Data Collection Method In Wireless Sensor Network Based On Compressive Sensing

With the wide application of Wireless Sensor Networks in military defense,environmental monitoring,medical diagnosis,intelligent transportation and other fields,the research of using the WSN to achieve the real-time monitoring and collecting the target signal has aslo attracted much attention.Wireless Sensor Networks are composed of a large number of small size,low cost,and free-from-environmental-constraints sensor nodes,which are densely arranged in the sensing area.However,due to the limited resources of a single node and the huge amount of communication in the network,the node energy consumption is too fast and the network lifetime is not long.And energy consumption of the data communication between the nodes accounted for a largest proportion in the consumption of total energy of nodes,so reducing network traffic can achieve energy efficient data collection.The Compressive Sensing technology utlizes the characteristic of temporal spatial correlation of the data adjacent node collect over time to reduce acquisition and transmission of the redundant data.This not only can reduce network traffic,but also can balance the load of communication.However,ordinary Compressive Sensing technology causes the high communication load of nodes in the early phase.Therefore,Hybrid Compressed Sensing proposed compressed sensing technology is used in nodes for processing data whose traffic is higher than the bottleneck,otherwise nodes transmit the original data.In order to reduce the cluster sensing network transmissions and prolong network lifetime,this paper proposed a data collection scheme based on the Hybrid Compressive Sensing(CS)technique for clustering Wireless Sensor Network(WSN).The scheme first divided the sensing area into several clusters according to the geographical location,assumed a virtual cluster head existed in the center of each cluster area,and selected the nodes which were one hop distance from the virtual cluster head as candidate cluster head.Second,a Minimum Spanning Tree(MST),which chose sink as root node and connected each virtual cluster head,was generated by the Prim algorithm.Third,starting from the sink,it chose cluster heads from candidate cluster head for clusters in each branch of the MST using Dynamic Programming.Finally,a backbone tree that connected all CHs to the sink in thesequence of MST was constructed.Simulations confirm that,when the compressive ratio is 10,compared with clustering without CS? SPT without CS?SPT with hybrid CS ? and clustering with hybrid CS,the reduction ratio of traffic of our method respectively are 65%?55%?40% and 10%.