The Research Of Clustering Routing Based On Particle Swarm Optimization For Wireless Sensor Networks

Wireless sensor network (WSN) is greatly promoted by the mature ofsensor technology, network communication technology and embeddedtechnology. And it is widely used in national defense military, agriculturemonitoring, environment monitoring, space exploration, industrial monitoring,intelligent transportation, intelligent household, health care and other fields. Inthe monitoring area, it collects and processes data, then send the data to theobserver.Wireless sensor network is consist of large number of nodes in the form ofself-organization according to the efficient routing protocol. So routing protocolis important to ensure high quality of communication and high energy utilizationin wireless sensor network. The clustering routing not only has high scalability,but also is high efficient in terms of energy and data transfer rate. But after aperiod of time of the operation, it is difficult to balance the residual energy of allnodes in the network. And with the increase of number of nodes, routing searchspace increase exponentially. So it is important to design an efficient path searchalgorithm to solve those problems. This article studies the clustering routing, and then put forward clusteringalgorithm based on improved particle swarm optimization and routing protocolbetween clusters. At first, several classical routing protocols and particle swarmoptimization are analyzed. Then the inertia weight of particle swarmoptimization is improved to make particle swarm optimization converge quicklyat the early stage of the search, and in the later stage, make particle swarmoptimization search accurately. And the proposed algorithm is applied toclustering routing protocol for wireless sensor network. At last, a new fitnessfunction and the mathematical model of particle are proposed. And new clustercompetition mechanism is introduced. Fitness function contains energyconsumption of communication between nodes, the remaining energy of clusterheads, distance between cluster heads and normal nodes within the clusters, andenergy-consuming balance of nodes. In the process of iterations of the algorithm,optimal cluster heads are selected by comparing the fitness values of particles.The simulation of the proposed algorithm NDAPSO-C by Matlab shows that theproposed algorithm cluster more uniform, balanced energy consumption toavoid network hole, prolong the lifetime of network, and make base stationreceive more data compared with LEACH、LEACH-C and PSO-C. After thecompletion of the clustering, in order to solve the low energy efficiency, one hoprouting and multiple hops routing are combined to transmit data betweenclusters. The next hop cluster head is selected based on the values of path costwhich is consist of distance between clusters and base station, energy of cluster nodes and the number of nodes in the clusters. The simulation of the proposedalgorithm CRBED by Matlab shows the proposed algorithm can balance energyconsumption, make base station receive more data and prolong the lifetime ofnetwork compared with LEACH-C and EEUC.