The Imaging Method Of Forward-looking Radar Based On GPU

The maximum difference between forward-looking imaging and traditional radar imaging is that the imaging scene is located ahead of the radar line of sight.So the quality of forward-looking imaging is entirely determined by the azimuth and elevation two-dimensional angle resolution of the radar system.In general,a large aperture array antenna structure must be used to obtain enough azimuth and elevation two-dimensional angle resolution.However,the aperture of antenna is usually limited for most radar systems such as missile system,airborne system,and so on.This does not only greatly limit the number of antennas but also reduce the angle resolution of space detection.Therefore,it is of crucial practical value to study the moving platform forward-looking radar imaging method to improve the angle resolution and breakthrough the limitation of antenna aperture.In order to realize the high resolution imaging of forward-looking radar,the phased array echo modeling method based on random radiation field is proposed in this paper.The method requires each array to adopt random amplitude modulation and phase modulation to change the shape of spatial pattern.In the multi-cycle,through the random amplitude modulation and phase modulation to achieve the transmit signal is non-coherent in space.This can form multiple phase centers which will enhance the inconsistent of far-field wavefront information and increase the difference of radiation information in the spatial distribution,so that the target within the same beam are identified with a greater possibility.We process the echo data by pulse compression and speed compensation before imaging,which corrects the envelope migration of echo data and ensures the sufficient signal to noise ratio during imaging.In order to ensure the reconstruction performance of forward-looking imaging,this paper proposes a forward-looking imaging algorithm based on the image total variation constraint with reference to the two step iterative threshold optimization method.The experimental simulation results show that the proposed method is effective.In addition,in order to improve the speed of forward-looking imaging algorithm,this paper studies the parallel algorithm based on GPU architecture on the basis of the in-depth analysis of forward-looking imaging algorithm.The process of algorithm design combines the characteristics of forward-looking imaging to improve the observation process,solving optimization and other parts.Parallel development on the GPU platform is implemented,which solves the problem of slow processing of forward-looking radar imaging to a certain extent.The simulation experiment demonstrates the validity and feasibility of the proposed method.