Research Of Multi-channel Radar Forward-looking Super Resolution Imaging Method

By placing multiple channels perpendicular to the track of the platform,the multichannel radar can realize two-dimensional imaging ahead of radar,which has significant application value in national defense construction and economic development fields such as missile guidance,sea and air monitoring and autonomous navigation.However,the length of the platform and other factors determine that the aperture length is usually small,which results in a great impact on the azimuth resolution,thus bringing a lot of inconvenience to the application of the multi-channel radar,so the implementation of multi-channel radar forward-looking super resolution imaging is of great significance.In this paper,the research work is carried out on the forward-looking superresolution imaging of multi-channel radar.Firstly,the echo expression of multi-channel radar is deduced according to its spatial geometry and working mode,then its azimuth resolution and the point spread function are analyzed.On this basis,the imaging convolution model is established,a multi-channel radar forward-looking super-resolution imaging method based on the accelerated iterative shrinkage threshold algorithm is studied,and a radar image enhancement method based on the convolutional neural network is proposed.The main contents are as follows:1.According to the spatial geometry and working mode of multi-channel radar,the echo signal model is established,the characteristics of multi-channel radar echo are analyzed,and the Doppler parameter characteristics and azimuth resolution of multichannel radar are quantitatively analyzed.and the imaging convolution model is established,which provides a theoretical basis for the follow-up research of superresolution imaging algorithm.2.A multi-channel radar forward-looking super-resolution method based on accelerated iterative shrinkage threshold algorithm is studied.Iterative threshold shrinkage algorithm through gradient optimization algorithm of iterative operation solving inverse convolution model,however the algorithm has low convergence efficiency.By introducing the vector extrapolation based on Taylor expansion to accelerate the algorithm,and taking the second order vector extrapolation method to accelerate iterative shrinkage threshold algorithm,this algorithm can effectively realize the multi-channel radar forward-looking super-resolution imaging.3.A radar image enhancement method based on convolutional neural network is studied and proposed.In view of the disadvantages of existing convolutional network,such as insufficient feature propagation and difficult information transmission between front and rear convolutional layers,the method adds weighted dense connection to the nonlinear mapping layer and the features of each layer are input to the subsequent layer,which is helpful to reduce the gradient loss of the designed network,take the advantages of residual learning and feature fusion,strengthen feature propagation and feature reuse,and then improve the image enhancement performance.The above research work has been checked by MATLAB simulation tools.The results show that the research work in this paper can realize multi-channel radar forward-looking super-resolution imaging and radar image enhancement.