Research On Node Localization Of Water Environment Monitoring Network For The Three Gorges Reservoir Area

Three Gorges Project has attracted worldwide attention and occupied an importantposition in China’s national economic development. Three Gorges reservoir waterenvironmental safety issues has been highly concerned. The existing monitoringequipment, methods and means can not meet demand of monitoring tasks. Ability ofenvironmental monitoring in Three Gorges reservoir is needs to be improved urgently.Wireless sensor networks which have features as low-cost, small size, low power andself-organizing can be used in Three Gorges Reservoir Water Environment Monitoring,which facilitate continuous, real-time monitoring in reservoir and early warning. Formonitoring applications, in order to accurately determine the possible sources ofpollution, data must be provided with nodes position information. Therefore, nodelocalization problems of large-scale wireless network in Three Gorges reservoir waterenvironment monitoring were studied in this paper.Large-scale wireless sensor network for monitoring in Three Gorges reservoir hasobvious banded distribution characteristics, which has caused great difficulties to nodeslocalization and often there are special area distribution networks as C and Y type.Special complex environmental factors were rarely considered in existing localizationalgorithm for large-scale regional networks. Based on actual needs of project, withtheoretical analysis and simulation experiment and combined with massive bandeddistribution characteristics of wireless sensor networks in Three Gorges Reservoir area,a localization algorithm based on game theory was proposed. Main work as follows:①Sensor network distribution characteristics of Three Gorges Reservoir wasIn-depth analysis, and anisotropy of complex regional networks was researched. Alocalization method based on game theory was proposed. Detailed analysis, modelsolution and concrete algorithm steps of the algorithm were given.②Using MATLAB simulation platform to verify localization algorithm based ongame theory in complex distribution area, and to compare with several algorithms.Experimental results show that the algorithm has achieved good localization effect inboth isotropic and anisotropic networks.③In the application environment, value of node localization error is difficult toobtain. With statistical sampling inference theory, model based on anchor nodes toachieve specific network node localization error estimation was proposed. Effecting ofmodel was analysed through experiments on multiple networks. Results show that the model can effectively estimate node localization error of network.