Research On Target Localization And Tracking Based On Wireless Sensor Network

Wireless sensor network is a new kind of emerging information acquisition and processing technology with the development of wireless communication technology,embedded technology,sensor technology,and distributed information processing technology.It has a very broad application prospects in the military,industry and social life,and location-based services are the foundation of all applications.Therefore,wireless target positioning and tracking are one of the key technologies in wireless sensor network,and it has important research significance.Based on the existing wireless positioning and tracking methods,this dissertation studies the target wireless positioning and tracking technology to improve the location accuracy,reduce the computational complexity and energy consumption.The main research contents and achievements are as follows:1.None of the existing positioning algorithms can adapt to all the localization scenes,especially in some specific scenes.(1)For the observation node is uniform circular array distribution and the target near the center of the circular array,there is a problem that the estimation error of the traditional positioning algorithm becomes larger due to the inversion of the singular matrix,this dissertation proposes a discrete variable constrained least square method based on hybrid measurement model(TDOA/TOA).The algorithm avoids the inverse operation of the singular matrix and improves the positioning accuracy.In addition,the algorithm's data processing technology can be extended to other positioning models;(2)to deal with the low positioning accuracy and large computational complexity of the traditional algorithms for positioning multiple dynamic targets,this dissertation presents a modified least square method.The algorithm considers the global constraint condition,introduces the Lagrangian multiplier technique and the quasi-Newton's BFGS iterative formula,which avoids the calculation of Hessian matrix,reduces the computational complexity and improves the positioning accuracy;(3)to solve the problem that the estimation error of the traditional algorithm is greatly increased due to the target near the reference site or any axis,this dissertation presents optimized two-stage least square method.The algorithm avoids the fatal flaw in the traditional algorithms by selecting the reference site again and rotating coordinate system,the increase of the computational cost of the algorithm improves the positioning performance of the algorithm.2.To cope with the problem that the traditional algorithms have a large estimation errors caused by NLOS measurement error,this dissertation presents an optimized localization algorithm for identifying and attenuating NLOS propagation error,which uses the Bayesian sequential decision to identify NLOS propagation,and proposes modified KF to smooth the measured data to reduce the NLOS effect.Finally,the residual weighting algorithm is used to estimate the position of target.The algorithm can effectively reduce the estimation error caused by the NLOS effect;the physical and simulation experiment verify the feasibility of the proposed algorithm.3.To solve the problem of different types of measurement data fusion under heterogeneous network,this dissertation proposes a conformal OCKF positioning and tracking algorithm.The algorithm introduces the additional variables to represent the nonlinear terms in the state vector,and proposes the adaptive weighting factor to adjust the specific gravity of the different types of signal input systems,and provide a reliable guarantee for the design of high precision algorithms.Then,the moving target is tracked by OCKF to update the state vector,velocity vector and additional variables.The simulation results show that the proposed algorithm has superior performance and robustness.4.To deal with the computational complexity and high energy consumption problems caused by large amount of data collection,this dissertation presents target positioning and tracking optimization algorithm based on compression perception.This thesis designs a sensing matrix that satisfies the RIP,and presents a sparse adaptive orthogonal matching pursuit algorithm to estimate the number of targets,and the target position vectors can be accurately recovered.The experiments results show that the proposed approach can locate the multi-target accurately under the condition of less data acquisition.