Millimeter-Wave Image Reconstruction Algorithm Based on Minimum Entropy for 2-D MIMO Array

The 2-D multiple-input–multiple-output (MIMO) array is an effective millimeter-wave (MMW) imaging scheme. However, the spacing of the 2-D real aperture MIMO array transceivers cannot be too small due to hardware-level limitations. This means that it is difficult to satisfy Nyquist’s sampling law in the azimuth direction, i.e., the array is sparse, which leads to extremely high sidelobe energy. Conventional fast algorithms based on the wavenumber domain are difficult to solve this problem. Although compressed sensing (CS) can reconstruct more information, it is not practical due to its excessive computational burden. To obtain high-quality 3-D images, this article proposes an MMW imaging algorithm based on the minimum entropy. In our scheme, the range scaling algorithm (RSA)-based forward and backward sensing operators are constructed in conjunction with the previously mentioned MIMO-RSA imaging method, which avoids large-scale inversion of the matrices. In addition, this article proposes an alternating direction method of multipliers (ADMM) in combination with RSA to solve the imaging problem with the expectation of obtaining an image with the minimum entropy. Meanwhile, the iterative solution process and the selection of various parameters are given using image entropy (IE) as the evaluation index. Finally, the simulation and experimental results verify that the algorithm can acquire high-quality images efficiently.