Research And Implementation Of Passive Bistatic Radar Target Detection Method Based On TACAN Signal

The passive bistatic radar has the advantages of strong anti-interference ability,silent detection,and not easy to be destroyed by the enemy,and has attracted wide attention.The TACAN signal has a high space-to-air emission power.When the TACAN signal is selected as the radiation source of the passive bistatic radar,it is beneficial to the detection of highaltitude and long-range targets by the passive bistatic radar.At the same time,TACAN base stations are widely distributed in coastal areas,which is conducive to the construction of the passive bistatic radar depth early warning detection network,and to improve the military strength of China's coastal defense,which is of great significance to national air defense security.This thesis takes TACAN signal as the irradiation source to study the method and implementation of radar target detection.The main work is summarized as follows: 1.There are two high-amplitude sidelobes in the distance dimension of the TACAN signal ambiguity function.To solve this problem,this thesis uses sidelobe identification and mismatch filtering to deal with the sidelobes.The TACAN signal has good recognition ability in Doppler dimensionality,and the distance unit between the sidelobes and the main peak is fixed.Based on this characteristic,this thesis proposes a method for sidelobe recognition.This method is used to find the main peak by comparing the values of different distance units of the same Doppler unit.After the main peak is found,the characteristics of the fixed distance unit between the main peak and the sidelobes are used to identify the sidelobes.Different from the traditional matched filtering method,the cost function of the mismatch filtering method is composed of two parts,one part is the loss of signal-to-noise ratio of the main peak after mismatch,the other part is the energy accumulated by the energy of the sidelobes.Besides,this method adds a modulation factor to the cost function to balance the signal-to-noise ratio loss of the main peak and sidelobes suppression effect.Simulation experiments show that both methods have obvious suppression effect on the sidelobes.2.In the realization process of target detection,the calculation amount is very large.In order to meet the requirements of real-time signal processing,GPU+CPU heterogeneous architecture is used for parallel acceleration calculations.In the clutter suppression and mismatch filtering methods,there are steps of autocorrelation matrix and cross-correlation matrix inversion.This step requires a large amount of calculation,and there is no library function in the existing database that can be called.This thesis chooses Gaussian elimination method for matrix inversion.The inversion process uses parallel elimination to realize parallel computing.In the process of distance-Doppler,the amount of calculation is large when performing fast Fourier transform.To solve this problem,this thesis adopts the method of simultaneous filtering and decimation.This method not only reduces the amount of calculation but also resists spectrum aliasing.The simulation experiment compares the running time of each module of the signal processing flow.Compared with the CPU,the running time of the architecture selected in this thesis is significantly shortened,indicating that the parallel architecture has good acceleration performance.