Study On The Modeling And Imaging Algorithm Of Forward-looking Scanning Radar

Radar has all-weather,all-day,long-range target detection and positioning capabilities.Radar high resolution forward imaging,which refers to microwave high resolution imaging of the front area of the aircraft,has always been a cutting-edge technical problem in battlefield reconnaissance,target surveillance,precision guidance and other important military fields.In recent years,this technology has been widely concerned by domestic and foreign research institutions.The process of forward-looking scanning radar imaging can be regarded as a process of solving an inverse problem.The inverse problem can be modeled and mathematically solved to improve the azimuth resolution of forward-looking radar imaging.Since the relative phase and the off-grid effects of the targets are not considered,the existing non-coherent echo model and its corresponding super-resolution algorithms are still insufficient.Therefore,based on the forward-scanning radar geometry model and electromagnetic wave propagation effect,a reasonable signal model should be proposed.Based on this,solving the inverse problem is of important theoretical significance and application value for improving the performance of forward-looking imaging.This paper focuses on the establishment of forward-looking scanning radar echo model and the algorithm of target angle estimation.The main contents are as follows:1.According to the geometric relationship of the moving platform,the non-coherent echo model of the forward-looking radar is derived.Based on this,an interpolation iterative adaptive algorithm(IIAA)is proposed.This method uses the low-pass characteristic of antenna pattern to suppress the high-frequency noise through low-pass filtering,and recovers the azimuth signal that is not acquired due to low pulse repetition frequency by interpolation.Under the same signal-to-noise ratio(SNR),the resolution performance is obviously better than the existing traditional algorithms.In the real data processing,the resolution is improved by 4 times compared with the real aperture imaging,and the SNR is significantly improved.2.The relative phase influences the resolvability of the non-coherent model.This paper reveals the reason why phase shifts affect deconvolution and proposes a coherent signal model by replacing the antenna power pattern with the radiation pattern.On this basis,this paper proposes a spectrum truncation preprocessing technique(STP),which can greatly improve the computational efficiency before frequency domain super-resolution.Compared with the traditional time domain super-resolution method,the time-consuming can be reduced by up to two orders of magnitude,making real-time super-resolution imaging feasible.3.Forward-looking scanning radar usually increases the pulse repetition frequency(PRF)in order to obtain the azimuth angle of the scene target more accurately.However,due to the ambiguous system distance and the revisit rate,it is impossible to increase the PRF without limit,and this method will bring the calculation of exponential growth.In this paper,an off-grid model based on the linear approximation of the time domain antenna pattern is proposed.This model can break the limitation of PRF and estimate the angle of the target deviating from the sampling interval to reduce the estimation error.Based on this model,this paper proposes an off-grid iterative self-decision approach(OGISA),which can automatically determine the target number and iteratively estimate the target angle.Under the same SNR condition,the performance of the algorithm is better than that of off-grid sparse Bayesian learning(OGSBL)and off-grid orthogonal matching pursuit(OGOMP),and its effectiveness is verified by real data processing.