Research On Precise Node Localization Algorithm For Three-dimensional Wireless Sensor Network

Node Localization is one of the key technologies for three-dimensional wireless sensor network(WSN).The accuracy of node location is crucial in most WSN applications.To improve the practicability and effectiveness of algorithm,two novel three-dimensional localization algorithms are proposed based on the two localization requirements of static anchor node and mobile anchor node localization in this dissertation,which provides a new idea for the research of three-dimensional localization.DV-Hop(Distance Vector-Hop)algorithm is a range-free localization algorithm.Due to its simple implementation,DV-Hop algorithm has become one of the more widely used algorithms at present.However,the research of DV-Hop algorithm in three-dimensional space is not mature.Based on the error analysis of DV-Hop algorithm,the improved 3D localization algorithm based on hop-count and differential evolution(DEHDV-Hop)is proposed to reduce the localization error.Firstly,the definition of the continuous hop-count is introduced according to the relationship between the number of neighbor nodes,and the upper bound of hop-count under the definition is proved.The global discrete hop-counts are converted to continuous values through node information broadcasting.Secondly,based on the error caused by actual and estimated distance of anchor nodes,this dissertation introduces concept of trust degree to updates the average hop distance of unknown node.The estimated distances between nodes are calculated by using new hop-counts and average hop distances.Finally,the hop-counts are segmented to reconstructed a new fitness function.The differential evolution algorithm is used to iteratively get the locations of unknown nodes.The simulation results show that DEHDV-Hop algorithm has lower localization error than comparison algorithms under various parameters.Path planning model has always been a research hotspot based on mobile anchor node localization.In this dissertation,an incremental localization algorithm based on mobile anchor node path planning(ICMAPP)is proposed considering static path planning.Firstly,the unknown nodes obtain the distances.The locations of the virtual anchor nodes with shortest distance from unknown nodes to the linear moving path of anchor node is calculated by extreme learning machine and the method of mathematical.That is,the abscesses of the unknown nodes are determined.Secondly,ICMAPP selects the first located node in the light of sequence of location and the anchor node moves to it.When it moves to the nearby,trilateral localization method is used to locate it accurately.By projecting the location of the first located node,the approximate locations of the subsequent unknown nodes are estimated by using the ranging error and the mobile anchor node moves to it.Two or more non coplanar virtual anchor node locations are selected to locate the unknown node.Similarly,ICMAPP uses incremental calculation to locate the remaining unknown nodes in turn.Finally,the three-demensional space is divided into several layers,and the whole area is located according to the above method.The ICMAPP algorithm has lower localization error and higher coverage,which is tested by simulation experiments.