Signal Processing Of Waveform Agility Radar And Implementation Based On GPU

The working environment of radars is becoming more and more complex,and it is of great significance to improve the anti-jamming capability of radars in complex electromagnetic environments.The traditional pulse Doppler radar with fixed parameters is easy to be intercepted.The enemy's jamming equipment can quickly analyze the signal parameters of the intercepted radar and generate the corresponding jamming signal,which poses a great threat to the radar detection performance.Waveform agility radar has excellent anti-jamming performance because of its random agility of carrier frequency,repetition frequency and coding parameters,which is not easy to be intercepted by the enemy.However,the nonuniform sampling of waveform agility radar in the slow time dimension makes the traditional signal processing method based on coherent assumptions no longer applicable.Based on this,this paper studies the signal processing method of waveform agility radar.The main work contents are as follows:1.Aiming at the problem that the traditional radar signal processing method is no longer applicable when the parameters of radar pulse signal such as carrier frequency,repetition frequency and coding information are jointly agile,the signal processing method of waveform agility radar is studied.First,the signal model of waveform agility radar is studied,and its ambiguity function is introduced with chirp and phase-coded signals as examples,and the range resolution,speed resolution and anti-jamming performance of waveform agility radar are analyzed.Then,the signal processing flow of traditional radar is introduced in detail.Finally,the signal processing flow of waveform agility radar is given,and the effectiveness of the method is verified by simulation.2.Aiming at the clutter suppression problem of waveform agility radar,four clutter suppression methods are studied.1)Clutter suppression method based on pulse train cancellation.This method takes the whole pulse train as the unit for clutter cancellation processing,which will lead to a smaller first blind speed,and is suitable for the clutter suppression problem of low-speed moving targets in the ground clutter environment.2)Clutter suppression method based on clutter reconstruction.Firstly,the clutter signal matrix is reconstructed according to the clutter peak points of the original range-velocity twodimensional matrix and the amplitude distribution of the clutter points in the Doppler dimension.Then,the goal of filtering the clutter is achieved by subtracting the clutter signal matrix from the original signal matrix.This method has a good suppression effect on static clutter.3)Clutter suppression method based on FIR filter.According to the clutter spectrum characteristics of the clutter signal in Doppler dimension,the corresponding filter is designed to filter out the clutter.This method can suppress the clutter signal and keep the signal of the doppler channel of interest undistorted.4)Clutter suppression method based on generalized matched filter.Firstly,training samples are selected from the echo data to estimate the covariance matrix of the cell under test.Then,the covariance matrix is used to whiten the echo signal to suppress clutter.This method is suitable for clutter suppression in a stable environment.3.The GPU parallel implementation method of waveform agility radar signal processing algorithm is studied.Adopt the idea of modular programming in software engineering,the algorithm flow is divided into pulse train matched filter module,array error correction module,integrated beamforming module,clutter suppression module and target detection module,etc.,according to the realization function.Also,the function and resource allocation strategy of each module is introduced in detail.The processing results of the measured data of MIMO radar show that each module has a higher acceleration ratio than the CPU serial calculation method,and the acceleration effect is good.