Localization Algorithm Using Particle Swarm Optimization In Wireless Sensor Network

Wireless Sensor Network (WSN) has the characteristics of low cost, low powerconsumption and self-organization etc. It is widely used in industry, military, agriculture,medical and many other fields. The information of node localization in WSN is theprecondition of all applications. The monitoring data is unvalued without the correspondinglocation information. The traditional measurement technology such as Received SignalStrength Indicator (RSSI) has the disadvantage of poor positioning accuracy. Therefore, theintelligent computing algorithms are introduced into WSN to solve the problem of nodelocalization gradually.In the field of intelligent localization algorithm, the Particle Swarm Optimization iswidely used as the reason of high positioning accuracy and low complexity, etc. Based on thestudy of Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Multi-ObjectiveParticle Swarm Optimization (MOPSO), three node localization methods are proposed withcombining the RSSI measurement technology and the intelligent algorithm in this paper.(1) The genetic localization algorithm based on reliance degree and geometric constraint.In the stage of selecting unlocalized node, choose the high positioning accuracy with reliancedegree. In the stage of genetic localization, constraint the initial population with the method ofgeometric constraint in order to reduce the searching space and speed up the convergencespeed. Use the real number coding as the coding operator and roulette algorithm combiningwith elite-preserving strategy as the selecting operator. The arithmetic crossover and theuniform mutation are adopted as the crossover operator and mutation operator respectively.(2) The chaos particle swarm optimization localization algorithm based on level-basedreliance degree scheme and premature detection. The process of localization is divided intotwo stages. In the first stage, select the unlocalized node depends on the positioning sequenceby level-based reliance degree scheme and calculating the distance between the unlocalizednode and the neighborhood anchor with RSSI measurement techniques. In the second stage,the particle swarm premature is avoided by chaos disturbance and chaos disturbance is set offby premature detection.(3) The localization algorithm based on multi-objective particle swarm optimization.Based on multi-objective particle swarm optimization, a localization algorithm is proposed tosolve the multi-objective optimization localization issues in wireless sensor networks. Themulti-objective functions consist of the space distance constraint and the geometric topologyconstraint. The optimal solution is found by multi-objective particle swarm optimizationalgorithm. Dynamic method is adopted to maintain the archive in order to limit the size ofarchive, and the global optimum is obtained according to the proportion of selection.The simulation results show that: the genetic localization algorithm based on reliancedegree and geometric constraint can estimate the position of the unknown nodes with lessanchor nodes and improve the positioning accuracy efficiently. The chaos particle swarmoptimization localization algorithm based on level-based reliance degree scheme andpremature detection can improve the localization accuracy and settle the problem of theparticle swarm premature effectively. The localization algorithm based on multi-objective particle swarm optimization can find the optimal solution and also achieve the betterpositioning accuracy and convergence rate.