In The Wireless Sensor Network Location Technology Research Based On Distance Measurement

As an emerging technology, wireless sensor networks have the potential to be widely applied in many fields of environmental monitoring, industrial control, smart grid and health care, for its advantages of large scale, self-organization, flexible, low power, low cost, etc.As one of the key technologies of wireless sensor networks, sensor nodes localization technology is the basis of providing the position information of the monitoring events. This thesis focuses on the problem of localization in wireless sensor networks. For the purpose of improving the accuracy of the localization algorithms, this thesis has studied the problems of static node locating and mobile node locating, and designed and implemented the RSS-based localization application system.The main research work and achievements are in the following areas:1) The weighted centroid localization algorithm based on RSSThis localization algorithm converts RSS to distance and utilizes weighted centroid algorithm to estimate the position of the blind nodes. When noise variance is4dB and the number of reference nodes is5, the localization accuracy increases by46.33%than the centroid localization algorithm.2) The enhanced alternating combination trilateration localization algorithm based on TDOAThis algorithm improves the localization accuracy of the ACT algorithm by considering the triangular shape of three reference nodes and the estimated position of the "vertex" relative to the triangle, and utilizes weighted centroid algorithm to estimate the position of the blind nodes. When ranging noise variance is30%of the true distance and the number of reference nodes is7, the localization accuracy of the algorithm increases by25.23%than mutilateration algorithm and12.24%than ACT algorithm.3) The mobile node localization algorithm based EKPFThis algorithm is the improvement of PF, which uses the approximate posterior probability by updating from EKF instead of the one step state transition probability of PF as the important probability, therefore effectively improves the localization performance of the mobile nodes. With the parameters set in this article, the localization accuracy of EKPF increases by17.13%than PF. And we find that EKPF can effectively solve the problems of divergence of EKF filter by simulation.4) Design and implementation of the RSS-based localization application systemThis localization system uses ZigBee nodes of the laboratory, and utilizes msstatePAN protocol stack to form ad hoc network, and applies the weighted centroid localization algorithm based on RSS. The test results show that the absolute localization error is0.57m and the relative localization error is13.44%. Compared to the centroid localization algorithm, localization accuracy of this system increases by59.68%.