Information metasurfaces and intelligent metasurfaces

Abstract

Metamaterials and metasurfaces have inspired worldwide interest in the recent two decades due to their extraordinary performance in controlling material parameters and electromagnetic properties. However, most studies on metamaterials and metasurfaces are focused on manipulations of electromagnetic fields and waves, because of their analog natures. The concepts of digital coding and programmable metasurfaces proposed in 2014 have opened a new perspective to characterize and design metasurfaces in a digital way, and made it possible to control electromagnetic fields/waves and process digital information simultaneously, yielding the birth of a new direction of information metasurfaces. On the other hand, artificial intelligence (AI) has become more important in automatic designs of metasurfaces. In this review paper, we first show the intrinsic natures and advantages of information metasurfaces, including information operations, programmable and real-time control capabilities, and space–time-coding strategies. Then we introduce the recent advances in designing metasurfaces using AI technologies, and particularly discuss the close combinations of information metasurfaces and AI to generate intelligent metasurfaces. We present self-adaptively smart metasurfaces, AI-based intelligent imagers, microwave cameras, and programmable AI machines based on optical neural networks. Finally, we indicate the challenges, applications, and future directions of information and intelligent metasurfaces.