Study On The Clustering Of Large Scale WSN Based On Local Similarity Of Network Structure

Wireless sensor network(WSN)is a new type of self-organizing network,as a hot spot on the current international network WSN has received much attention.Compared with the traditional network,there are hundreds or even more of nodes in wireless sensor network and the nodes are widely distributed in the sensor filed.The energy supplied by the battery of nodes in wireless sensor networks is usually limited,so in order to prolong the network lifetime most researchers are trying to reduce energy consumption when in the process of designing routing algorithm or other strategy.With the increase of application requirements,the scale of wireless sensor network is increasing,and the limitation of energy,computing power and communication range of the nodes has led to the contradiction between the energy consumption and lifetime of network becoming more and more prominent.So in current research,how to improve the energy efficiency of the large-scale wireless sensor network become the hot topic and the optimization of clustering algorithm is one of the key technologies of improving energy efficiency.The routing algorithm can greatly influence the performance of the network,especially under the large-scale wireless sensor network,mass data need for a long-distance transmission,so the performance of the routing algorithm will directly affect the service lifetime of the network.In general,designing routing algorithm for wireless sensor network is affected by many factors,not only to reduce the energy consumption,but also balance the energy from the perspective of the whole network structure.Only in this way can extend the service lifetime of the whole network finally.According to the characteristics of the large-scale wireless sensor networks and the application requirements,this thesis analyzes the performance of LEACH in brief firstly.We has carried on the simulation analysis of its performance in 100 node scale,1600 node scale and 3000 node scale respectively,then analyzed its limitation when the size of network increased,which came to a conclusion that puts forward a better clutering algorithm to solve the problem of large scale network is particularly important;Then the link prediction of complex network was studied in detail in this thesis,on the basis of considering the energy,three commonly used kinds of Link Prediction algorithm based on local similarity of network structure are improved,and make it suitable for wireless sensor networks.Then on the basis of the three Link Prediction algorithm combined with LEACH clumping ideas,we put forward three kinds of WSN Clustering algorithm based on local similarity of network structure.They are recorded as CHBCN-E,CHBPA-E,CHBRA-E respectively,for convenience,this thesis referred to them as CHBLP(Clustering Hierarchy-based on Link Prediction).Finally,simulation experiments for these clustering algorithms under different scale scenario were carried out.At the same time in order to analysis the network performance this thesis compares them with the classical clustering algorithm-LEACH detailed.The experimental results show that the proposed CHBLP clustering algorithm makes the network has a good reliability,and at the same time it can reduce the network energy consumption and prolong network lifetime.