Research And Implementation Of Airborne Radar Super-resolution Imaging Method

Airborne radar super-resolution imaging has increased great interest in the fields of self-landing,terrain awareness and supply dropping.However,the current airborne radar super-resolution imaging methods still have the problem of low image resolution,and it is difficult to satisfy the requirement for engineering applications due to the large amount of computation of imaging algorithms.Therefore,research on airborne radar super-resolution imaging algorithm and CUDA paralleling implementation are of important value.Focusing on the problems of airborne radar super-resolution imaging,the thesis carries out research work on airborne radar forward-looking imaging echo model,super-resolution imaging algorithm based on total-variance and CUDA implementation of airborne radar imaging method.The details are as following:1.According to the geometry model of airborne radar forward-looking imaging,the range history between target and carrier platform is derived.The established echo model of airborne forward-looking radar and the established convolution signal model of airborne radar azimuth echo provide the theoretical basis for the researches on deconvolution methods.2.Aiming at the ill-posed problem of traditional deconvolution method,the Bayesian model is used to derive the iterative formula of the super-resolution imaging algorithm based on total-variance.The super-resolution imaging performance of the proposed method is verified by point target and surface target simulation.3.Aiming at the high computational complexity of airborne radar super-resolution imaging algorithm,the general parallel computing platform architecture and CUDA programming model are introduced.This thesis proposes the parallel implementation method of airborne radar super-resolution imaging algorithm.This design of the airborne radar super-resolution imaging software system architecture and the development of its various sub-modules are completed,which effectively improving the super-resolution imaging efficiency.The research and implementation of the above super-resolution imaging algorithms have been verified by simulation experiments.The results show that the proposed method can not only improve the resolution of airborne radar super-resolution imaging,but also satisfy the requirement for engineering applications.