A Clustering Algorithm For The Wireless Sensor Network Based On Improved PSO

Along With the progress of electronic information technology and network communication technology,which promote the development of Wireless Sensor Network(Wireless Sensor Network,WSN)rapidly,the application of Wireless Sensor Network(WSN)makes human life rich.The research about WSN involves many disciplines,which is a hot concern at home and abroad.WSN network combines many nodes into a self-organizing networks by a variety of routing protocol,after processing the data collected,they will be transported to the base station,and transmission medium is the sensor nodes.The nodes in a network are interdependent and coordinate with each other,after a period of using time,the WSN network will make the energy of the nodes in the network size is uneven,which is caused by the influence of environment and the data size,and the nodes will change along with the environment,such as temperature,the increase of the irregular touch attack,the irregular increase of the network node and dynamic topology changes,etc.,they will make route and energy consumption of WSN network exponentially increase,so as to shorten the network life.So the study of these problems are meaningful.This paper will research hierarchical routing protocol based on clustering,firstly analyzes various WSN routing protocol and particle swarm optimization(PSO)algorithm in detail,and then puts forward a clustering routing algorithm based on improved PSO algorithm(mPSO-W).The algorithm is improved mainly from the following two aspects.The first is to improve the clustering of inertia weight of the WSN,which is to enhance the global search ability at the beginning of the iteration and enhance the effect of local search in the late iterations,it also can avoid the trapping into the local optimal solution in advance.The second is to improve the fitness function,that considering the factors of influencing the WSN node clustering: The energy consumption standard deviation between clusters,energy consumption balance degree within a cluster,node communication distance within the cluster and energy ratio of the cluster head.A new fitness function can select cluster head better,which make the node clustering more even and prolong the life cycle of wireless sensor network(WSN).Finally,comparing with LEACH,SEP,WPSO algorithm,this paper puts forward the improved clustering algorithm called mPSO–W,which makes the wireless sensor network has the characteristics such as clustering evenly,balanced energy consumption and low data loss rate Etc.,at the same time,inhibits the generation of uneven energy distribution,so as to achieve the effect of the prolong the network life cycle.