Research On Nodes Deployment And Data Collection Technology In Wireless Sensor Network

Data collection is one of the key problems in Wireless Sensor Network (WSN). And thenodes deployment is the basis of data collection. Generally, the power supply of nodes islimited by battery power which makes it hard to maintain long-term network lifetime. And theincrease in the number of nodes will affect the stability of the network. It restricts the practicalapplication of WSN in various fields. The key of WSN data collection is how to prolong thenetwork lifetime. Therefore, this paper studied the key problems of nodes deployment anddata collection that includes clustering routing protocols, data fusion and so on. On the basisof the optimal nodes deployment, we proposed the clustering compressed sensing datacollection technique of wireless sensor network. The main research works are as follows:1) In order to reduce the cost and complexity of wireless sensor networks, we need toreduce the number of sensor nodes. However, it is difficult to ensure the completeness andaccuracy of the acquired data by using a small number of nodes. So this paper proposed anodes deployment algorithm based on kriging interpolation method in cluster. Firstly, we usethe regular grid method to pre-deploy nodes in each cluster of the whole area. Thenconsidering the scenario in which sensor readings are spatially correlated, we adopt thekriging interpolation to estimate the data of every node that is assumed to be reduced in turn.Lastly, the error of estimation can help to find the optimal nodes to be reduced. The resultsshow that this deployment algorithm needs fewer nodes with acquiring a sufficiently accurateapproximation of data, testing by the simulated and real world data.2) The distance-based particle swarm optimization clustering method was proposed toreduce the uneven energy consumption of WSN. Under the condition of energy, it considersthe relationship of distance from cluster head to the base station and cluster head to the nodedistance to update the target function. It makes that the closer distance from the node to thebase station and the other nodes the larger probability of node to become cluster head.Simulation results demonstrate that the proposed algorithm can balance the energyconsumption and prolong the network lifetime effectively.3) The more compressed raw data are, the smaller node data transmission is, and thelower energy consumption is, which is beneficial to prolong the network lifetime. But thehigh-compression tends be bad for the accurate data reconstruction. A novel data collectionmethod of WSN based on clustering compressed sensing was presented to solve thecontradiction between data accuracy collected and energy consumption in sensor nodes,which considers the sparsity of the monitoring signal in wireless sensor network. In the proposed method, the (Low Energy Adaptive Clustering Hierarchy) LEACH protocol wasadapted to select cluster head and cluster formation from the random arrangement of sensornodes, and the Gaussian random matrix was utilized to linearly compress sensor data in thecluster by every cluster head. Then the compressed information was transmitted to the basestation. It reduces data transmission and energy consumption, thus improving the lifetime ofnetwork. According to monitor signal’s regional smoothness of sensor nodes, the differentialtransformation regularization was adopted to reconstruct receiving linear compressionprojection information by the base station. Simulation experiments show that the datacollection method of WSN based on clustering compressed sensing can guarantee dataaccuracy collected and improve the lifetime of whole network simultaneously.