Assessment and Optimization of Disordered Distribution Reflecting Metasurfaces for Computational Imaging

Microwave computational imaging (MCI) provides an alternative method to realize fast-speed, high-resolution imaging with low system complexity and cost. However, the MCI scheme needs a large amount of independent measurement modes to acquire the scene information, which puts a burden on the antenna design. The adoption of metasurfaces, especially the reflecting metasurface (RM) architectures, provides a new idea to generate low-correlated radiation patterns that can be used as quasi-orthogonal measurement modes in the MCI. Nevertheless, most studies in the aperture design aspect focus on designing the metamaterial element itself and analyses about the element distribution are neglected. Therefore, in this paper, we investigate the influence of element distributions on the performance of the RM. Firstly, a series of full wave simulations about the metamaterial element topology are conducted. Then, the compact disordered distribution is generated according to the simulated annealing (SA) algorithm, which is used to arrange the metamaterial elements. Finally, based on CST simulations, the singular values (SVs) of measurement modes generated by the metasurface are calculated, which can be used to evaluate the performance of the RM. According to the simulated results, a conclusion can be drawn that the element distribution form would influence the quality of measurement modes generated by the RM.