Study On Passive Localization Method With The Minimum Number Of Observers

Passive target localization systems with multiple stations use multiple sensors to passively receive radiation signals to achieve target location,which have widely used in wireless sensor networks,radar and navigation.However,there is a strong nonlinearity between the measurements and the target parameters.The traditional closedform solution algorithms and their improved method introduce auxiliary variables to construct a pseudo-linear localization equation to improve the target location performance.In three-dimensional space,the accurate localization can be obtained with at least five observers.Thus the problem of four-station passive localization is a challenge for the multi-station passive localization.In addition,multi-static passive location problems with unknown location of the transmitter,sensor position error and clock biases make the multi-static passive location in uncertain environment difficulty.This paper focused on the four-station passive localization problem and multi-static passive localization problem with unknown transmitter,sensor error and clock biases.The main work in this paper is as follows:(1)Existing localization method with minimum number of observers has poor localization accuracy in high level noise.Aiming at the problem,a new two-step weighted least square with Newton’s method(TSWLS-NM)is proposed in this paper.Firstly,a new pseudo-linear location equation is established.Secondly,according to the relationship between the auxiliary variable and the target position,a quadratic polynomial about the auxiliary variable is constructed,which is solved by the Newton iteration method.Then the estimated value of the target position is obtained and is further corrected.Finally,the theoretical analysis and simulation verification are carried out.The experimental results show that the TSWLS-NM algorithm can realize the target localization in the case of four stations.The proposed method can achieve the CramerRao Lower Bound(CRLB)under moderate noise.(2)When there is relative motion between the target and the observer,a TSWLSNM algorithm based on frequency difference of arrival(FDOA)and time difference of arrival(TDOA)is proposed in this paper based on the work(1).Firstly,a new pseudolinearized moving target location equation is established.Secondly,a quadratic polynomial about the auxiliary variables of position and velocity is constructed,which is solved by Newton iteration method.Then the joint estimated value of the target position and velocity is obtained and is further corrected.Finally,the theoretical analysis and simulation verification are carried out.The experimental results show that the TSWLS-NM algorithm can solve the moving target localization when using four moving stations.The proposed method has better robustness than the existing methods.(3)Aiming at the multi-static passive localization problem with unknown transmitter,sensor error and clock biases,a joint location algorithm based on bistatic range and direct range is proposed in this paper.Firstly,new auxiliary variables are introduced by using the algebraic characteristics of the measurement and the target position,and a joint estimation location model of the target position,transmitter position,offset and auxiliary variables is constructed.Secondly,the weight is designed to obtain the joint weighted least squares(WLS)estimate.Thirdly,the correlation between auxiliary variables and unknown parameters is considered to further improve the target localization accuracy.Finally,the theoretical analysis and simulation verification are carried out.The experimental results show that the proposed algorithm can accurately solve the target position in the strong nonlinear scenario.The proposed method is a closed solution algorithm,which is more suitable for practical application scenarios.