Research On Passive Time Difference Of Arrival Location Based On Cross-correlation

Passive time difference positioning technology has become a hot research field in recent years due to its high positioning stability and practicability.It has obvious anti-interference characteristics in actual signal source localization(such as electronic countermeasures).The time difference of arrival(TDOA)positioning technique is to construct a hyperbola by converting the time difference of the same signal to a plurality of monitoring stations,and to construct a hyperbolic curve by using a hyperbolic method.Among them,the cross-correlation algorithm is widely used because of its strong achievability and high stability.In this Master's Dissertations,the basic theory of passive time difference positioning is systematically expounded,introducing the time difference of arrival in passive time difference location from the nature of hyperbolic method,autocorrelation and cross-correlation function to the traditional cross-correlation method and second correlation method.The positioning has been introduced,and the algorithm and reasoning based on the cross-correlation time difference algorithm have been fully carried out.In this process,the fast calculation method and the problems in the cross-correlation algorithm are also explained.The cross-correlation passive time difference localization algorithm has low noise immunity and low estimation accuracy under low SNR conditions,including the noise received by the signal during propagation,the ability of the cross-correlation algorithm to analyze the signal,and the peak error caused by peak jitter when the peak value of the function peak.It is summarized into three stages: signal receiving end,signal processing end and peak value.The whole process is simulated by Matlab software.These three stages use the mainstream algorithm to improve it separately,improving the accuracy of the delay estimation.Among them,the signal denoising methods adopted at the signal receiving end are wavelet denoising and singular value decomposition.The wavelet denoising is obtained by wavelet transforming the received signal,and then inversely transforming the thresholded signal to obtain the denoised signal.The singular value decomposition is constructed by the specific form of the signal.The matrix is subjected to singular value decomposition,and the signal matrix is reconstructed for the singular value before twice the dominant frequency,and then the summed sum is obtained to obtain the denoised signal.The signal analysis methods used in the signal processing end are single weighted and double weighted.The weighting function selects the characteristics of the integrated two-way signal and has a smooth coherent transform window that suppresses the peak of the noise,sharpening cross-correlation function.Since the error of cross-correlation peak value may be generated due to the fast Fourier transform fence effect in the cross-correlation algorithm,the methods of smoothing the signal peak at the spectral peak are least squares fitting,cubic spline Interpolation,refine the peak byinterpolation and improve the accuracy of the value.The simulation experiment is carried out on the improved algorithm.The simulation results show that the improved algorithm has better delay estimation accuracy than the original algorithm under low SNR conditions.On the basis of the above work,a brief introduction is made to the monitoring equipment for signal reception,and then each of the three stages of signal receiving end,signal processing end and peak value selection is selected: sparse wavelet based noise reduction Fourier transform time delay estimation,TDOA delay estimation based on improved second correlation algorithm,and generalized cross-correlation time delay estimation algorithm based on cubic spline interpolation are used to verify the application.The data analysis results show that the algorithm is improved in the actual environment.Like the simulation experiment,it has better delay estimation performance than the original algorithm.