Deployment Optimization Strategy Of Wireless Sensor Network Nodes Based On Particle Swarm Optimization

Sensor node deployment is one of the fundamental issues in wireless sensor networks(WSN), which is the basis of the normal working of WSN and affects the performance andeffectiveness of the networks. Usually, Wireless sensor network works in extremely harshenvironments where human beings can not guard. Sensor nodes are deployed in the monitoringarea by random spilled. This deployment is prone to produce coverage holes and affects thecoverage rate in wireless sensor networks. In addition, wireless sensor networks are relied onbattery-powered with limited energy. Due to physical constraints, it's difficult to charge orreplace the battery. So, energy limit is one of the most critical constraints of the entire WSNdesign, which directly determines the life cycle of the network.To solve the above problems, this paper launches an in-depth study from two aspects,including improving network coverage rate and extending the network life cycle. This papermainly has done the following work:(1) This paper summarizes the research status of WSN and nodes deployment issues,carries out a systematic classification of the WSN nodes deployment strategy, analyzes severalexisting nodes deployment strategies, proposes the inadequacies and directions need to beimproved in the existing algorithms.(2) Through analysis and comparison of the existing self-organizing cluster-based routingalgorithms and the cluster head selection mechanism, this paper finds that classical LEACHalgorithm is unreasonable in the election of cluster head nodes, and proposes a PSO-basedcluster head selection mechanism. On this basis, an improved topology optimization strategybased on dynamic neighborhood particle swarm optimization (DNPSO) is proposed. Followingthe principle of the overall network energy minimization, and considering the impact ofdistance on the energy consumption in the data transfer process, the improved algorithm isbetter able to balance the energy burden in wireless sensor networks, and optimize the networktopology. Finally, through simulation experiments, it proves that the proposed algorithm is ableto balance nodes energy consumption in the network, and extend the network life cycle.(3) A new deployment strategy of wireless sensor network nodes based on improveddynamic multi-populations particle swarm optimization is proposed. Network coverage as thefitness function, the deployment of sensor nodes will be formalized as an objective optimizationproblem. By employing the k-means clustering algorithm, the population is divided into severalsub-populations, in addition, the population will be re-divided into new sub-populationsdynamically, which can weaken particles on the pursuit of local optima, effectively solve the "premature" problem of basic PSO algorithm, and accelerate astringent of the algorithm.Experimental results show that this deployment strategy can reduce the coverage holes inwireless sensor networks as much as possible and effectively improve the network coveragerate.