Non-Local Metasurface-aided Leaky-Wave Antennas

TL;DR

The paper tackles gain degradation in metasurface-aided terahertz leaky-wave antennas caused by angle-dependent incidence in local metasurfaces. It introduces a non-local metasurface framework described by four surface susceptibilities linked by GSTCs, enabling near-angle-insensitive beam steering around $2.5$ THz. A theoretical limit shows exact angular invariance is possible only for $ar{R}= extpm1$, and the authors implement a PSO-based optimization to realize near-insensitive performance with impedances $Z_1$, $Z_2$, and $Z_3$, then map these to physical unit cells using an iterative truncation approach. Full-wave validation in ANSYS HFSS demonstrates beam steering of roughly $-23^\circ$ to $+29^\circ$ over $2.0$–$2.7$ THz with stable gain and sidelobes, yielding more than a $3 imes$ improvement over conventional local metasurface designs. The approach provides a practical route to broadband, robust THz beam steering with holographic non-local metasurfaces.

Abstract

This work presents a non-local terahertz metasurface integrated into a leaky-wave antenna for robust, wide-angle beam steering. The metasurface encodes a holographic pattern by explicitly inducing tangential and normal susceptibilities, along with magnetoelectric coupling. This design maintains stable radiation performance even when the longitudinal wavenumber of the incident guided mode - and thus its effective impinging angle - varies as a function of frequency. In particular, we show that there exists a limit to achieving exact angular insensitivity and propose an optimization-based framework to obtain the required susceptibilities that closely approximate near angle-insensitive performance for stable beam-steering performance. Additionally, an iterative synthesis approach is introduced that maps abstract susceptibilities to physically realizable structures. Full-wave simulations demonstrate a beam-scanning range of nearly 50 degrees over the 2.0-2.7 THz band - a more than threefold improvement over conventional local-metasurface designs.

Figures (6)

• Figure 1: (a) Schematic of the proposed leaky-wave antenna and (b) the equivalent transmission-line model of the non-local metasurface.
• Figure 2: Angular and frequency dependence of the optimized unit cell. (a) Magnitude and (b) phase of the reflection coefficient as a function of incidence angle and frequency. (c, d) Comparison of the simulated reflection (c) magnitude and (d) phase against the target values at 2.5 THz.
• Figure 3: Comparison of the scattering parameters for Cell 1 and Cell 2 at the design frequency 2.5 THz: (a) reflection magnitude, (b) transmission magnitude, (c) reflection phase, and (d) transmission phase.
• Figure 4: (a) Schematic of simulation setup employing impedance boundaries that implement the optimized $\{Z_1,~Z_2,~Z_3\}$ for Cell 1 and Cell 2. Full-wave radiation patterns (gain) of the LWA employing (b) the proposed angle-insensitive metasurface over 1.9-2.6 THz and (c) local metasurface over 2.4-2.6 THz.
• Figure 5: Physical realization procedure of unit cells. (a) Ideal sheets in periodic boundary conditions. (b) Truncated impedance strips that possess similar angular responses. (c) A single layer separated from the multilayer stack and its equivalent physical pattern. (d) Complete unit cell constructed by stacking realized sheets. (e, f) Optimized geometries of the constituent impedance layers ($Z_1$, $Z_2$, and $Z_3$) arranged from left to right for (e) Cell 1 and (f) Cell 2. Note that the middle layer ($Z_2^{(2)}$) of Cell2 is implemented as a simple open circuit.