Study On Passive Localization Technology Using Multiple Sensors

Electromagnetic spectrum resources are one of the most critical combat resources in the world on the battlefield.The use and denial of the various bands of electromagnetic waves can generally include three aspects: communication,detection and electronic warfare.Passive detection of electromagnetic waves is one important aspect of detection and the precondition of electronic countermeasures.Passive detection can also be used as an auxiliary means of other active detection.So passive detection has important practical significance and value.Passive localization is one of the final processes and outputs of passive detection with advantages of a longer detection distance,security and stealth,etc.Passive localization can be situational awareness,interference guidance,and so on.Passive localization outputs can be used as guidance for active detection systems such as radar and assault weapons.However,the development of low intercept and low detection performance of radiation sources such as modern radar and communications poses a great challenge to passive detection and passive localization.Therefore,it is necessary to study the specific passive localization method for radiation source with low intercept and low detection characteristics on the basis of existing passive localization technology.The main research work of this dissertation is summarized below:For signals with low signal-to-noise ratio,the DOA measurements obtained by each observation station are less and the measurement error is large,so the passive localization performance is not ideal.Therefore,a new passive DOA localization method based on restricted constrained total least squares is proposed in the first part.First,the nonlinear DOA measurement equation is transformed into a linear equation.According to the first-order Taylor approximation of the linear equation coefficients,the linear mapping between the DOA measurement noise and the equation coefficient noise is obtained.Then the localization objective function is obtained based on the restricted constraint least squares algorithm,the approximate closed solutions of the localization results are obtained by deriving the partial order and ignoring the high order term of the noise,the regularization parameters of the restricted constraint least squares algorithm are determined by analyzing the bias and mean square error.Theoretical analysis and simulation experiments show that the algorithm has improved the accuracy of the passive DOA localization algorithm when the number of DOA measurements is small and the measurement error is high.When the signal-to-noise ratio is low,it is possible to obtain a higher accuracy by using the cross-correlation method to measure the Time Difference of Arrival(TDOA).The crosscorrelation method measures TODA requires that the radiation source signals intercepted by the different stations are retransmitted to the same location,such as the main observation station.In addition the observation station placed in the moving platform and other reasons leading to the observation station position error may affect the passive localization accuracy performance.Aiming at these two problems,a retransmitted TDOA localization algorithm based on constrained total least squares is proposed in the second part.Firstly,the nonlinear localization equation of retransmitted TDOA is transformed into a direct linear equation without intermediate variables,and then transformed into constrained optimization problem and unconstrained optimization problem based on the constrained total least squares algorithm,finally,the approximate closed solution is derived.The simulation results show that the algorithm has better performance than the relative algorithm when the position error of the observation station is large.For the centralized TDOA,each TDOA relates to the relative distance between the main station and the radiation source which the localization performance is limited by.The relative position between the station and the radiation source is made even by using the decentralized TDOA passive location.In the same station position arrangement,using the decentralized TDOA to passive locating the source can obtain more gently or even better positioning accuracy performance with the change of the radiation source position.Therefore,a passive localization method using decentralized TDOA is proposed in the third part.The linear equations of the intermediate variable are directly eliminated and transformed into an unconstrained optimization problem without considering the nonlinear constraint between the intermediate variable and the radiation source.Based on the constrained global least squares method,the BFGS method is used to solve the problem.And the unbiased and error variance of the proposed method are analyzed.The simulation results show that the root of the mean square error is better than that of the relative method when there is no position error in the observation station.When the location error of the observation station is small,the signal-to-noise ratio is low or the position of the radiation source is far from the observation station,the algorithm has better accuracy performance.Aiming at the bottlenecks in data transmission and calculation of the original Direct Position Determination(DPD),there are two decentralized parallel DPD methods using the TDOA and FDOA are proposed in the fourth part.The first method adopts the decentralized pairing scheme,and the signals intercepted by each observation station are only transmitted one by one between the stations.The data transmission and calculation are distributed parallel to calculate the cross ambiguity function(CAF)between the observation stations.The construction of the cross ambiguity matrix(CAM)only satisfies the full rank condition.The second method is based on the relation formula of any two CAFs.Using the reduction approach to parallel calculate the remaining CAF and complement the CAM at each observation station in a decentralized manner.Performance analysis shows that both methods can reduce the data transmission of the DPD method,reduce the computational complexity and improve the computational efficiency.The simulation results show that the accuracy of the two methods is better than that of the two-step method,and the two methods can achieve the ideal accuracy performance when the signal-to-noise ratio is low.When the signal-to-noise ratio is high,the second method and the original direct positioning method can achieve equal accuracy performance.