Machine Learning and Digital Twinning Enabled Liquid Crystals mm-Wave Reconfigurable Devices Design and Systems Operation

Machine learning is a versatile but little-understood approach in tunable RF electronics. This paper serves as the first to investigate the feasibility of taking inspiration from machine learning and the recent advances in digital twinning for designing liquid crystals-based reconfigurable devices in the regime of gigahertz (GHz) and terahertz (THz), targeting demanding applications in 5G/6G wireless backhaul links, satellite internet mobile terminals, and smart sensing. States of the arts are reviewed, with a prime design example of liquid crystals-based enclosed coplanar waveguide phase shifter illustrated, highlighting the opportunities offered and challenges encountered.