Research On Weighted Centroid Node Localization Algorithm In Wireless Sensor Networks

In Wireless Sensor Networks(WSNs),node location technology is one of the key technologies,which is consist of the range-based and non-range-based location algorithms.Based on the analysis of the status quo of WSNs positioning technology,this paper analyzes and evaluates the performance indicators and challenges of the algorithm for location algorithms based on ranging and non-ranging as well as their basic principles.The weighted centroid location algorithm based on non-ranging is emphatically studied,and the existing algorithm is improved.The main work and innovation points are as follows:(1)For the traditional weighted center-of-mass localization algorithm whose localization accuracy is affected by the number of anchor nodes and whether they are uniformly distributed,a weighted center-of-mass node localization algorithm based on MeanShift clustering is proposed.When an anchor node is selected,it is the first step to select a rectangular area in 3D space,while dividing the rectangular area into a 3D cell mesh,and then creating a target model;Secondly,the MeanShift algorithm is used in the target model to find dense and uniformly distributed anchor nodes by offset vector iterations;Finally,the coplanar degree is introduced to select the combination set of four non-coplanar anchor nodes for positioning.Simulation results show that the proposed algorithm has higher positioning accuracy compared with the centroid and weighted centroid localization algorithms when the number of anchor nodes,total number of nodes and communication radius are different.(2)To address the problems that the accuracy of the traditional weighted center-of-mass localization algorithm is not high and the received signal strength indication(RSSI)in the ranging stage is affected by the ambient noise,a weighted Beetle Antennae Search(BAS)based algorithm is proposed.Firstly,in the ranging stage,the obtained RSSI values are averaged as reference values,while the threshold value of the filtered data is set,and then the Gaussian filtering is performed;Secondly,the filtered values are used to weight the RSSI;Finally,the initial calculated positions are optimized to obtain better positioning accuracy by using the BAS for constructing the target search function.The simulation results show that the algorithm presented in this chapter has fewer iterations.When the number of anchor nodes and nodes is different,the positioning accuracy is higher.