Data Collection Based On Compressed Sensing For Wireless Sensor Networks

As a novel network,the WSN has been widely applied to numerous fields such as our daily life,industry as well as military and etc.The WSNs consist of large numbers of sensor nodes,who are able to communicate wirelessly.These nodes are widely distributed in the monitoring area to carry out data monitoring,data collection as well as data transmission.But the power of nodes is limited and it can't be supplemented timely,once the energy in the wireless sensor networks is used up,the performance of the whole networks will be affected or even closed to be paralysis.Therefore,it becomes a hot topic of how to reduce the overall energy costs on the process of data collection.On the basis of investigating the WSNs and data collection techniques,this thesis simply introduces the Data Collection Technology and meanwhile invokes the Compressed Sensing Technology.Addressing the energy-consumption problem when applying the traditional data collection mechanism,the data collection technology based on compressed sensing is presented by combining the compressed sensing with the data aggregation,and then the advantage of energy consumption is studied.Next,addressing the intensive projection problem of CS,we extend the exist sparse random projection mechism by appending the distance selection strategy into the process of data collection and come up with a compressed sensing algorithm based on shortest path tree,which can reduce the overall energy consumption by letting the interest nodes choose the projection node or sink node as the destination node according to distance selection strategy to transmit data packet on the process of data transmission.Finally,combining with routing strategy and data aggregation technique as well as differences between nodes,we propose the CN-MSTP,which selects different nodes randomly as projection nodes,and set every projection node as the root of the minimum spanning tree,then carry out data aggregation with the most possibility to select the interest node according to the greedy principle on the building of the minimum spanning tree.Soon afterwards,we propose the eCN-MSTP by putting the sink node into the construction of the spanning tree.Simulation results indicatie the proposed two algoriyhms have the advantages of saving the overall energy consumption as well as load balance and hence prolong the lifetime of the network.