Researches On Microwave Imaging Based On Compressed Sensing

Essentially,microwave imaging belongs to inverse scattering problems,which has a wide range of applications in civilian and military fields.In this dissertation,we mainly focus on microwave imaging based on compressed sensing.First,the wireless localization problem is investigated.By leveraging the antenna with random radiation characteristics,we propose a new approach for wireless localization based on meta-aperture and compressed sensing(CS).Utilizing the plasmonic dispersion of a magnetically uniaxial metamaterial,a pushpin-shaped autocorrelation characteristic of the localization system can be obtained to ensure high localization precision.Based on the CS,localization can be obtained with much less data and much faster data processing than conventional methods through solving the inverse scattering problem between scattering echoes and radiation fields.Simulation and experimental investigation are conducted,and the results demonstrate the effectiveness of the proposed approach.Second,the microwave imaging with random muti-beam illuminations is investigated.Utilizing the random radiation field obtained by phase-array with random phase,a single frequency microwave imaging method with optimal spectrum efficiency and imaging range is proposed.Based on compressed sensing,the simulation of the microwave imaging with random illuminations is investigated,which shows the potential of super-resolution imaging.In order to solve the matrix equation of imaging model,the key of imaging system with random illuminations is the produce of stochastic electromagnetic radiation field distribution,which makes the measurement in random way.Singular value decomposition and mutual incoherence function is adopted to calculate the row correlation and column correlation of the system functions respectively,which is applied in the imaging range optimization process.Simulation and experimental investigation are demonstrated,and the results prove the effectiveness of the proposed method.Last,the real-time microwave imaging is investigated,especially for objects behind walls or surrounded by closed obstacles.The imaging equations are derived based on a combination of the conventional model of inverse scattering problem and the concept of compressed sensing.We experimentally demonstrate highly efficient imaging for complex-structured objects surrounded by a closed obstacle.Making use of the spatial sparsity of objects and obstacles in the imaging region,the proposed imaging can be implemented using a time-division multi-antenna system with a reduced number of transmitting antennas.Without any notable loss of imaging quality,the imaging time can be reduced by two orders of magnitude compared with conventional inverse scattering methods,implying a potential to implement real-time imaging.According to the prior information of wall,a new imaging method based on compressed sensing under inhomogeneous background is also proposed.