Research On Sparse Reconstruction Technology For Microwave Staring Correlated Imaging Of Moving Target

In order to extract more information from scattering echo, improving the resolution of radar imaging is one of the most main method. The microwave staring correlated imaging, whose radiation field randomly changes with time and space, can obtain high transverse resolution that breaking antenna aperture diffraction limit. The microwave staring correlated imaging method for moving target is a new imaging system to obtain super-resolution characteristic of moving target. But the correlation processing method for static target is not suitable for moving target. In the aerial staring irradiated area of ground-based radar, the moving target is often sparse distributed. In this paper, we focus on the research of sparse reconstruction technique for moving target based on microwave staring correlated imaging theory by applying the characteristic that the scatterers of the moving target are sparse distributed.Firstly, the influence of target motion on microwave staring correlated imaging is analyzed. Based on the model of microwave staring correlated imaging for stationary target, the model for moving target of microwave staring correlated imaging is established from the physical process. Then combined with the prior information that the scatterers of moving target are sparse distributed, classic sparse reconstruction algorithm, such as Orthogonal Matching Pursuit(OMP), Sparse Bayesian Learning(SBL) and Focal Underdetermined System Solver(FOCUSS), is applied to the system of microwave staring correlated imaging for moving target. A series of simulations are carried out to analyze the target motion influence on applying the classic sparse reconstruction to microwave staring correlated imaging.Secondly, the sparse imaging algorithm for moving target in microwave staring correlated imaging is studied. In the imaging scene of moving target, the classic sparse reconstruction algorithm will fail if ignoring the effect of target motion. Therefore, this paper puts forward two sparse imaging algorithms for moving target in microwave staring correlated imaging.(1)For frequency random hopping system, the property that the frequency of transmitted signal is random hopping is used for velocity pre-estimated and the motion compensation for radiation field matrix is carried out. Then, the target image will be got through correlated process.(2)From the view of Bayesian statistical optimization, we modify the classic SBL algorithm and obtain the target image and velocity through joint estimation. The simulation results show that the proposed algorithm can obtain precise velocity estimation result and high resolution image at the same time.Finally, the problem of grid mismatch in the microwave staring correlated imaging for moving target is researched. When applying the classic sparse reconstruction algorithm to microwave staring correlated imaging, no matter how fine the meshing, the grid mismatch problem will still be there. In order to solve this problem, this paper puts forward two imaging algorithms.(1)For frequency random hopping system, the velocity of moving target is pre-estimated and motion compensation is performed. Then by using the characteristic that the grid positions and the grid biases have the same sparse structure, the classic OMP algorithm is modified by adding the step of grid mismatch error correction.(2)Converting moving target imaging problem to the estimation of continuous parameter, this paper uses the prior information of the target and designs the optimization function to solve the location of scatterers directly. This method avoids the dependence of grid in classic sparse reconstruction algorithm and obtains precise sparse reconstruction result when where is off-grid error through the simulation experiment.