Bayesian Compressive Sensing In SAR Imaging

Radar imaging is the use of electromagnetic waves as a detection signal,with highresolution imaging technology to observe the area or scene for high-precision twodimensional imaging technology.Since the method of matching filtering is limited by the Shannon-Nyquist sampling theorem,this method is difficult to break through the radar imaging performance.Compressive sensing is a new sampling theory proposed in recent years.The theory uses the a priori information of sparse signal to reconstruct the original signal by non-linear method through a small number of sampling signals,breaking the limitation of Shannon-Nyquist sampling theorem.Compressive sensing applications in radar imaging not only can greatly reduce the sampling rate,but also enhance the performance of radar imaging.In the case of low signal-to-noise ratio,however,the quality of the compressive sensing signal will be degraded,which is a typical problem to be solved.In addition,in complex-valued electromagnetic environment,the performance of radar imaging will be greatly affected especially in the interference environment.In the face of complex-valued electromagnetic environment,how to improve the performance of radar imaging has always been an urgent need to solve.Based on the Bayesian estimation theory,this paper discusses the complex-valued Bayesian compressive sensing reconstruction method and the Bayesian compressive sensing imaging strategy under the interference.This paper mainly from the following aspects:In this paper,the advantages and disadvantages of the compressive sensing reconstruction algorithm are studied,and some problems in the field of synthetic aperture imaging are discussed.This paper first studies the applicable scenarios and compares the advantages and disadvantages of various Bayesian compressive sensing reconstruction algorithms.The paper completes the theoretical derivation of the Bayesian compressionaware reconstruction algorithm and analyzes the complexity of each algorithm.For the problem of reconstructing the signal quality degradation in the case of low signal to noise ratio(SNR),this paper deeply studies the compressive sensing imaging algorithm based on Bayesian estimation theory,and proposes an improved complexvalued Bayesian compressive sensing reconstruction algorithm.The algorithm uses the phase information of the scene to divide the original complex-valued measurement equation into real and imaginary parts,and combines the imaging results of the real part and the imaginary part to improve the reconstruction result.Compared with the existing complex Bayesian compressive sensing reconstruction algorithm,the algorithm has low computational complexity,fast convergence speed and excellent imaging performance.Simulation experiments and measured data show that the improved complex-valued Bayesian compressive sensing reconstruction algorithm has a good reconstruction effect on the signal under low SNR.Facing the problem of how to improve the performance of radar imaging in complex electromagnetic environment,this paper proposes a Bayesian compression sensing SAR frequency combination optimization algorithm.The algorithm mainly includes two parts.First,a new frequency combination strategy is proposed.From the single Gaussian model specification step size to the mixed Gaussian model specification step size,the system frequency is quickly jumped out of the "hardest hit area" in order to reduce the number of iterations.Followed by the frequency combination selection problem,the algorithm draws on the Metropolis criterion,does not blindly reject the suboptimal solution,but accept the suboptimal solution with a certain probability to expect it to jump out of the local optimal solution.Therefore,the algorithm converges as much as possible to the global better solution.The simulation results verify the effectiveness of the algorithm.