Study On Localization Algorithm For Wireless Sensor Networks

Wireless sensor networks involve sensor technology, MEMS technology, distributed informationprocessing technology, wireless communication technology and many other subjects, and havebecome one of the hottest IT fields. Wireless sensor networks have advantages such as self-organizing,wide coverage and high fault tolerance, and also have features like low cost, as well as flexibility andconvenience in building and using. So wireless sensor networks are developing as a new manner ofaccessing information, and are widely applied in both military and civilian fields. As wireless sensornetworks continue to develop and mature, the application will become more widespread.Because the precise location of the node itself is the premise to provide location information formonitoring events, the node localization technology is a core supporting technology in wireless sensornetworks. The study of node localization techniques has a crucial theoretical significance and appliedvalue for improving the overall performance of wireless sensor networks and ensuring the reliability.But so far, localization accuracy has not yet reached a satisfactory level, and especially in the presenceof interference, it has yet to improve.This paper first discusses the basic principles of localization technology in wireless sensornetworks, classification methods of localization algorithms, and performance evaluation. Then,according to the defect that traditional RSSI (Received Signal Strength Indicator) localizationalgorithms are not accurate under non-ideal environment, we propose a new node-locating algorithmcalled CLSSM (Collaborative Localization Scheme From Spring Model), which is based on RSSIvalidation and spring model optimization. In order to improve the localization accuracy, the CLSSMalgorithm establishes a spring model for wireless sensor networks, and uses the connectivity-basedspring model to optimize the result of RSSI calibration algorithm. Simulation results show that underthe same experimental conditions, CLSSM algorithm is superior to the traditional RSSI locationalgorithms. CLSSM algorithm improves about19%in the localization accuracy, compared with RSSIcalibration algorithm.In addition, this paper delves into the localization technology for mobile nodes, and proposes anew localization method named Enhanced Monte-Carlo localization boxed. Based on the MCB(Monte Carlo Localization Boxed), EMCB introduces the crossover and mutation operations ofGenetic Algorithm into sample selecting, in order to make samples move towards regions with largevalue of posterior density distribution. So the distribution of samples is optimized, and the problem oflow sampling efficiency is solved. Simulation results show that compared with the MCB, the newalgorithm’s localization accuracy is improved by about17%, while the computational cost is muchsmaller.