Comparison of Reflectarray vs. Transmitarray Reconfigurable Architectures for 6G Sub-THz

Abstract

This paper compares reconfigurable reflectarray (RA) and transmitarray (TA) architectures for 6G sub-THz (D/F-band) applications, with emphasis on high-speed beam steering, aperture efficiency, and system integrability. We present an evaluation framework that integrates full EM co-simulation and link-level modeling to assess key figures of merit: η_ap, scan loss, fractional bandwidth, SLL, polarimetry, beam squint, and EIRP/G/T in backhaul and directional access scenarios. Three tuning mechanisms—2-bit MEMS, continuous liquid-crystal (LC), and varactor/CMOS—are discussed in terms of insertion loss, linearity, switching latency, bias requirements, and fabrication tolerances. The analysis shows that RA offers a simple feed but is susceptible to blockage and mutual coupling over wide scans, while TA minimizes blockage and facilitates multilevel true-time-delay for controlling squint, at the cost of multilevel assembly complexity. Case studies of sub-arrays at 130–170 GHz illustrate practical trade-offs between phase quantization (1–2 bit vs. continuous), dielectric loss, and bias routing versus wide-scan performance. This framework yields target-based architecture selection guidelines: RA for low cost and profile at moderate scan rates, TA for wide scan rates/high EIRP and advanced radio integration, and scalable unit-cell and control network design recommendations toward 6G reconfigurable metasurface antennas.