Research On Clustering Technology For Wireless Sensor Networks

In wireless sensor networks, the energy constraint of sensor node determines the lifetime of the network directly. Thus, the primary design goal of wireless sensor networks is making use of energy efficiently. In the process of data transmission of sensor network, the fusion of data first to reduce redundant data, and then forward the fusion data can greatly reduce energy consumption. This paper studied the clustering and routing techniques for wireless sensor network comprehensive and in-depth. The study of specific research and implementation of this work summarized as follows:(i) This paper proposed a double cluster head structure data collection protocol EADC based on clustering. When choosing the main cluster head, the node sends the main cluster-head competition message according to the residual energy and distance of its neighbor nodes. Then a vice cluster head is added in the cluster structure to responsible for forwarding data of the main cluster head. Vice cluster head nodes forward the data according to the layer structure.(ii) We implemented the LEACH, EADEEG, EADC protocol on OMNeT++ simulation platform, then compared and analysised the results of the simulation. The performance in EADC has increased significantly over LEACH, EADEEG, and the network lifetime is effectively prolonged.(iii) We proposed a routing quality protocol which based on k-means algorithm and the least square method. We isolated error data using k-means algorithm before the data fusion, and then used the least square method to fuse the remaining reliable data. This method improved the accuracy of fusion results is higher.