Research On Weighted Centroid Node Localization Algorithm Based On Kalman Filtering

The Wireless Sensor Network(WSN)has now developed into the fourth generation.It is low-cost and can provide regional information for people under different environmental conditions,therefore the application range of WSN has expanded from the initial military field to many other fields,such as waters,bridges,medical treatment and agriculture.Its network terminal is a large number of intelligent sensors,The human society senses,collects,stores,processes and transmits the occurrence of local areas through sensor nodes to complete the monitoring of the physical world.In each field of WSN application,node location information is indispensable,and improving the positioning accuracy of nodes has always been a research focus.At present,there are a large number of positioning algorithms,but the same algorithm cannot be used in different scenarios.This paper mainly focuses on the indoor environment,researches and analyzes the WSN node positioning,and establishes the simulation environment through MATLAB2016 software,The performance of positioning accuracy is simulated and analyzed.When the WSN is actually used in indoor environments such as factories,classrooms,warehouses,etc,there are many obstacles,and the ambient temperature and humidity are different,which may lead to a great difference between the RSSI(Received Signal Strength Indicator)of multiple groups from the same node measured by the unknown node.The inaccuracy of the RSSI measurement directly leads to the difference between the calculated distance and the actual distance,the error of the distance between the nodes will affect the positioning accuracy of the node position.In this paper,combining with the ranging model,the filtering performance of different filtering methods are analyzed.The Kalman algorithm is used to smooth the RSSI value.The simulation results show that Kalman filtering has higher accuracy and can achieve better filtering effect with fewer iterations.The Heron formula is used to optimize the weighted centroid localization algorithm(WCLA)in the localization algorithm.First of all,based on the distance between the node values to select optimal weighting factor coefficient to improve the weight factor,The selection of weight factors is simulated by setting the different of nodes distribution to verify its performance.Then the heron algorithm is used to optimize the selection of anchor nodes in the centroid localization algorithm(CLA),the first weighted centroid is used to obtain the virtual anchor nodes,and then the weighted centroid of the virtual anchor nodes is used as the unknown node coordinates.Finally,the H-WCLA algorithm,the traditional weighted centroid algorithm and the C-WCLA are compared by simulation,the simulation results show that the H-WCLA localization algorithm has good robustness,low computational power requirements,and improved node positioning accuracy.