Research And Improvement Of WSN Location Algorithm Based On Simulated Annealing

Node localization is an important supporting technology in the practical application of wireless sensor networks.The main work of this paper is to study the non-ranging localization algorithm,and improve the DV-Hop algorithm based on simulated annealing algorithm.The specific work is as follows:(1)The application and development status of wireless sensor networks are studied,and the method of node localization is mainly studied.Since the non-ranging node localization has relatively low hardware requirements,this paper mainly studies the non-ranging node localization algorithm,and simulates and analyzes several classical non-ranging algorithms,and summarizes the factors affecting the positioning performance.(2)Aiming at the problem that DV-Hop algorithm has a large average hop error obtained by each unknown node in the algorithm due to uneven node distribution,a simulated annealing localization algorithm based on ranging correction is proposed.For the case that the number of neighbor beacon nodes of unknown nodes is relatively large,the algorithm introduces the concept of near degree,which is used to correct the ranging value of the average hops from unknown nodes to neighbor beacon nodes,and then directly calculate the unknown nodes by least squares method;For the case where the neighbor beacon nodes are sparse,the weighted average hops are used instead of the near degree to correct the estimated distance,and the initial coordinates are obtained by the least squares method.Then the idea of simulated annealing is introduced to correct the initial coordinates.The simulation results show that the proposed simulated annealing localization algorithm based on DV-Hop ranging correction can achieve higher positioning accuracy.(3)Aiming at the problem of poor scalability and large localization error of DV-Hop algorithm in large-scale wireless sensor networks,a simulated annealing localization algorithm based on region-based grid scanning is proposed.The algorithm divides the region by Tyson polygon(Voronoi diagram)to reduce the energy loss and Mac conflict caused by network flooding in traditional algorithms.Then,using the estimated distance and hop count of the unknown node to the beacon node obtained in the previous step as constraints,the grid scanning method is used to narrow the area of the unknown node,and finally the simulated annealing algorithm is used to find the coordinates of the unknown node.The simulation results show that the algorithm can achieve better performance in scalability and positioning accuracy.