Joint Bias Registration And Target Localization Algorithm For Multi-transmitter And Multi-receiver Passive Radar

The multi-transmitter and multi-receiver passive radar does not actively transmit signals.Parameters such as time delay,angle and Doppler frequency difference are extracted from the transmitter signals which scatter from target and direct wave signals,and the position and velocity of the target are estimated based on these parameters.The above estimation requires that the measurement parameters are accurate and unbiased.However,in practical applications,the existence of fixed bias such as propagation congestion under multipath and clock bias will seriously affect the performance of target localization.Due to the strong nonlinearity and fixed bias of multi-transmitter and multi-receiver radar measurements,the existing localization algorithms cannot be directly applied in.In this regard,this thesis studies the multi-transmitter and multireceiver passive radar localization with fixed bias.Under different radar localization scenarios and estimation algorithms,an effective localization algorithm is proposed according to the algebraic characteristics of each localization equation.The main work of this thesis is as follows:1.Aiming at the problem of fixed bias in the Bistatic Range(BR)measurement of multi-transmitter and multi-receiver passive radar system,the Cramer-Rao Lower Bound(CRLB)performance under two estimation methods is compared.In this regard,a joint BR localization and error registration algorithm based on grouped Constrained Total Least Squares(CTLS)is proposed.The constraint-based CTLS model is established and solved iteratively by Newton method.On this basis,correlation processing is used to improve the target localization accuracy furtherly.Finally,the proposed algorithm is theoretically analyzed and simulated,and the localization performance under different estimation algorithms is compared.The influence of nearfield/far-field target and different bias levels on the localization and bias estimation accuracy is discussed.Simulation results show that the proposed group CTLS algorithm has better localization performance than the Two-stage Weighted Least Squares(2WLS)algorithm and the group Constrained Weighted Least Squares(CWLS)algorithm.2.In order to avoid the degradation of localization performance which caused by grouping optimization,a multi-stage weighted least squares(MWLS)joint BR localization and bias registration algorithm without grouping is proposed considering the BR localization scenario of multi-transmitter and multi-receiver with fixed bias.The first step is to construct the auxiliary variable to linearize the BR measurement equation,and obtain the preliminary estimation results of the target position and the bias according to the iterative weighted least square method.Then,a two-step dependent least squares algorithm is used to eliminate the influence of the correlation between the target location and the auxiliary variables,the unbiased target location estimation is obtained through error registration and posterior iteration,so as to improve the accuracy of target location estimation furtherly.Finally,the theoretical error analysis and emulation proof of the algorithm are carried out,the estimated performance under different algorithms is compared,and the influence of near-field target and far-field target,different bias conditions and different auxiliary variable selection on the localization accuracy is analyzed.The theoretical performance analysis and simulation results show that the localization estimation of the proposed MWLS algorithm can reach CRLB,and the performance is better than that of group CTLS,group 2WLS and group CWLS algorithms.3.Aiming at the problem of joint BR and direction of arrival(DOA)localization of multi-transmitter and multi-receiver radar under clock bias,a joint BR/DOA localization and error registration algorithm based on MWLS is proposed.Considering the influence of clock bias,the new auxiliary variables are selected to construct the pseudo-linear estimation model of BR/DOA joint estimation,and the iterative weighted least squares algorithm is used to obtain the initial estimation.On this basis,the correlation estimation processing and bias registration are improved.Finally,the theoretical analysis and simulation verification of the porposed algorithm are carried out.Compared with the group 2WLS and the posterior 2WLS algorithms,the proposed MWLS algorithm has better estimation performance of clock bias and target position,and the localization accuracy can reach CRLB.The CRLB of joint BR/DOA localization under bias is superior to the pure BR localization,which proves that the multi-domain information complementarity can improve the cooperative localization performance of multi-transmitter and multi-receiver passive radar system.4.The multi-transmitter and multi-receiver passive radar localization system uses multiple sets of bistatic range and Bistatic Range Rate(BRR)measurements to locate multiple moving targets.The existence of system bias makes the localization performance degrade,so this chapter proposes a BR/BRR joint bias registration and target localization algorithm based on MWLS.Firstly,according to the characteristics of moving target measurement information,a new auxiliary variable is constructed to pseudo-linearize the nonlinear BR/BRR joint measurement equation,and the joint estimation equation of target state and system bias is established.Then the improved correlation least squares algorithm is designed by using the constraint relationship between auxiliary variables and target position and velocity,and the system bias in BR/BRR measurement information is corrected to obtain the unbiased estimation of moving target state.Finally,the posterior iterative correction error is used to improve the estimation accuracy of target position and velocity furtherly.Theoretical analysis and emulation proof indicate that the algorithm of this chapter has good global convergence and the target localization performance reaches CRLB.Compared with group 2WLS and posterior 2WLS algorithms,the proposed MWLS algorithm has better performance in system bias and target position/velocity estimation.