Research On Microwave Correlated Sparse Imaging In Weak Randomness Of Temporal-Spatial Radiation Field

In recent years,there is an increasing demand for staring high-resolution imaging in key areas using static radar platforms.The traditional real-aperture radar technology generally uses the large-aperture antenna to generate a narrow beam.If the angular resolution of the real-aperture radar is improved,the size of the actual antenna needs to be increased.Due to the limitation of the actual antenna aperture,the real-aperture radar can not meet the need for high-resolution imaging.Microwave correlated imaging is a new type of radar imaging method based on temporal-spatial two-dimensional random radiation field and correlated processing method.It can achieve high-resolution imaging of the scene and provide a new idea to break through the limitation of real-aperture radar resolution.The performance of current correlated imaging methods has high requirements on the randomness of the temporal-spatial two-dimensional radiation field.However,due to the limitation of signal bandwidth and the number of elements,the randomness of the temporal-spatial radiation field is weak,which seriously affects the results of microwave correlated imaging.In order to solve this problem,the thesis improves the randomness of the temporal-spatial radiation field through the design of specific waveform.On the other hand,sparse reconstruction algorithms are proposed in the thesis,which can get better reconstruction results in the weak randomness of temporal-spatial radiation field.The main work of this thesis is as follows:In this thesis,the frequency hopping coding method of transmitting signal is designed to realize the specific waveform,so as to improve the randomness of the temporal-spatial radiation field.First,a construction method of the temporal-spatial radiation field based on random frequency hopping is introduced.The effects of signal bandwidth and frequency hopping interval on the randomness of the temporal-spatial radiation field are studied.Numerical simulations and results analysis are performed from the perspectives of the coherence,effective rank,and condition number of the radiation field matrix.Then,a construction method of temporal-spatial radiation field based on cyclic Costas frequency hopping coding is proposed.The results show that the randomness of temporal-spatial radiation field based on cyclic Costas frequency hopping coding is slightly stronger.In order to further enhance the randomness of the temporal-spatial radiation field,a construction method of temporal-spatial radiation field based on the frequency hopping coding optimized by annealing algorithm is proposed.Simulation comparisons and result analysis are carried out.In this thesis,the sparse reconstruction algorithm in the weak randomness of temporal-spatial radiation field is studied.Considering that the compressive sensing model is naturally consistent with the microwave correlated imaging model,and the target scene is generally sparse,the compressive sensing sparse reconstruction algorithm is used to reconstruct the target.Several different sparse reconstruction algorithms are compared and analyzed by simulation.The simulation results show that in the case of weak randomness of temporal-spatial radiation fields,the results obtained by general sparse reconstruction algorithms are not ideal.A sparse reconstruction method based on quantum-behaved particle swarm optimization is proposed to solve this problem.The method can search the optimal solution of the problem in parallel.The corresponding numerical simulation and analysis are performed.It can be seen from the results that the proposed method can not only obtain an ideal reconstruction result,but also has a certain tolerance to noise.Finally,in view of the slow convergence speed of the sparse reconstruction method based on quantum-behaved particle swarm,a calculation speed improvement method is proposed.The method uses the estimated support set to determine the approximate range of non-zero values in the sparse solution,so as to avoid unnecessary learning of algorithms and accelerate the convergence speed of the original algorithm.Simulation comparison and analysis are carried out.It can be seen from the results that the proposed method not only greatly improves the convergence speed of the original algorithm,but also obtains the same results as the original algorithm.