Research On Node Localization Algorithm For 3D Wireless Sensor Networks

Wireless sensor networks(WSN) node localization algorithm is a method to calculate the coordinates of unknown nodes rely on anchor nodes. Most of the existing researches of node localization are in the plane. In fact, since the environment of node localization algorithm in three-dimensional is complex, localization algorithms existing in two-dimensional plane are difficult to directly applied to the three-dimensional environment. Therefore, the node localization of WSN in three-dimensional space needs new theories and methods. Based on the theory of fuzzy geometric information, we introduce knowledge of mobile anchor node, Delaunay triangulation, adaptive node clustering and other knowledge research three new models to improve node positioning accuracy. Main work is as follows:On the basis of the problem of trilateral positioning for WSN node positioning accuracy, proposes a linear geometry node localization algorithm based on fuzzy theory(Mobile Anchor Node Localization on Fuzzy Information,MANLFI). The algorithm introduced knowledge of mobile anchor node with a view to calculate the initial coordinates of unknown nodes by measuring the angle between anchor nodes and unknown nodes. By weighting method, we calculated the final coordinates of nodes and updated the speed and direction of anchor nodes after each round. The simulation results confirm that the fuzzy information nodes localization algorithm based on mobile anchor nodes increases the utilization of anchor nodes, improves the accuracy of localization algorithms and effectively extends the network life cycle.After introducing the relevant knowledge of Delaunay triangulation, we proposed a node localization algorithm based on simple Delaunay triangulation and fuzzy information(Fuzzy Information Node Localization on Delaunay Triangulation,FINL-DT). Firstly, the algorithm achieves a simple Delaunay triangulation for anchor nodes in the network, realizing node localization by measuring direction angles and pitch angles between anchor nodes and unknown nodes in a triangle. After each round,this algorithm judge invalid anchor nodes and update invalid anchor nodes' position in network. Moreover, located nodes act as secondary anchor nodes to locate other nodes.The simulation results show FINL-DT algorithm improves the efficiency and accuracy of positioning, and lower the energy consumption of the network.This paper introduces the knowledge of node adaptive clustering, and proposes a node localization algorithm based on adaptive clustering and fuzzy information inthree-dimensional space. Initially, this method introduces virtual dimension to achieve dimensionality reduction and then starts adapting clustering for nodes in the same plane.And then, it achieves node localization with fuzzy information among anchor nodes and unknown nodes and improves the positioning accuracy by Taylor series. Simulation results show the algorithm improves the positioning accuracy and lowers the energy consumption of networks.