Harnessing Selective State Space Models to Enhance Semianalytical Design of Fabrication-Ready Multilayered Huygens' Metasurfaces: Part I - Field-based Semianalytical Synthesis

Imagine you are trying to build a magical window that can bend light (or radio waves) exactly how you want it to—focusing it like a magnifying glass, steering it like a laser pointer, or splitting it like a prism. In the world of physics, these "magic windows" are called metasurfaces.

For a long time, designing these windows was like trying to bake a perfect cake by guessing the ingredients. You'd have to mix the batter (design the structure), bake it (run a massive computer simulation), taste it, realize it's too salty, and then start all over again. This process was slow, expensive, and required supercomputers.

This paper, Part I of a two-part series, introduces a new way to bake that cake: a semi-analytical recipe book that tells you exactly what ingredients to use without needing to guess or run thousands of simulations.

Here is the breakdown of how they did it, using simple analogies:

1. The Problem: The "Black Box" of Design

Usually, to design these metasurfaces, engineers use "Full-Wave Optimization." Think of this as trying to find the perfect combination of spices for a soup by tasting every single possible mix. It works, but it takes forever.

The Issue: These metasurfaces are made of many thin layers of metal and plastic (like a sandwich). The waves bounce back and forth between these layers in complex ways. Ignoring these bounces leads to a bad design; accounting for them with traditional math is too hard; accounting for them with simulations is too slow.

2. The Solution: The "LAYERS" Recipe Book

The authors created a tool called LAYERS. Instead of guessing, they used a rigorous mathematical model (like a precise physics equation) to predict how the waves behave.

The Analogy: Imagine you have a complex machine with 5 gears stacked on top of each other. Instead of building the machine and spinning the gears to see what happens, you have a perfect blueprint that tells you exactly how fast the top gear will spin if you turn the bottom one.
The Innovation: They took a model that was originally used for simple, sparse wires and upgraded it to handle Jerusalem Crosses (a specific shape of metal pattern that looks like a cross with smaller crosses on the ends). They also made it work for dual-polarization, meaning the window works for waves coming in from any angle or orientation, not just one specific direction.

3. The "Lookup Table" (The Cheat Sheet)

The core of their method is creating a Lookup Table (LUT).

How it works: They ran their math model to find thousands of different metal patterns. For each pattern, they recorded:
1. How much of the signal gets through (Transparency).
2. How much the signal is delayed or shifted (Phase).
The Result: They created a massive "menu" where, if you want a signal to shift by exactly 45 degrees, you can instantly look up the exact metal pattern (the "recipe") that does that.
Why it's cool: Before this, finding that pattern might take hours of simulation. Now, you just look it up in the book.

4. The "Tuning Knob" (Leg Lengths)

The metal patterns they use are called Jerusalem Crosses. The authors discovered that the "legs" of these crosses act like tuning knobs.

The Analogy: Think of a guitar string. If you shorten the string, the pitch goes up. Similarly, if you change the length of the metal "legs" on the cross, you change how the wave behaves.
By simply changing the length of these legs in their math model, they could tune the wave to do exactly what they wanted.

5. The "Magic Trick" for Speed (Scaling)

One of the biggest hurdles in engineering is that a design that works at one frequency (like 20 GHz) might fail at another (like 22 GHz).

The Trick: The authors found a clever scaling rule. Instead of re-calculating everything for a new frequency, they realized they could just "shrink" or "stretch" the entire design mathematically.
The Analogy: It's like having a map of a city. If you want to know what the city looks like if you zoom in or out, you don't need to redraw the whole map; you just change the scale. This allows them to predict how the device will work across a wide range of frequencies instantly.

6. The Proof: The "Metalens"

To prove their method works, they designed a Metalens (a flat lens that focuses waves).

They used their Lookup Table to pick the right "recipe" for every tiny spot on the lens.
They built a virtual version of it and ran a final, high-precision simulation (the "taste test").
The Result: It worked perfectly! It focused the waves just like a real glass lens, but it was flat, thin, and made of standard circuit board materials.

7. What About Part II? (The AI Upgrade)

The authors admit that while their math is great, it's not perfect 100% of the time. Sometimes, the "recipe" is slightly off.

The Fix: In Part II of their series, they introduce an AI (Machine Learning) assistant.
The Analogy: Think of Part I as a master chef who knows 95% of the recipes perfectly. Part II is a sous-chef (the AI) who tastes the dish and makes tiny adjustments to get that final 5% of perfection. This combination allows them to design these devices almost instantly with near-perfect accuracy.

Summary

This paper presents a fast, accurate, and open-source toolkit for designing advanced "magic windows" for radio waves.

No more guessing: They replaced slow simulations with a smart math model.
Instant recipes: They created a lookup table so engineers can instantly find the right design.
Broadband ready: They figured out how to make these designs work across different frequencies.
Future-proof: This method sets the stage for an AI-powered design tool that will make building these devices as easy as downloading a file.

In short, they turned the difficult art of designing complex wave-manipulating devices into a simple, fast, and reliable engineering process.

Here is a detailed technical summary of the paper "Harnessing Selective State Space Models to Enhance Semianalytical Design of Fabrication-Ready Multilayered Huygens' Metasurfaces: Part I – Field-based Semianalytical Synthesis."

1. Problem Statement

Designing multilayered Huygens' Metasurfaces (HMSs) for microwave frequencies typically involves two stages:

Macroscopic Design: Determining the required surface impedance distributions to manipulate wavefronts (e.g., focusing, beam steering).
Microscopic Design: Finding specific physical meta-atom geometries (e.g., Jerusalem Cross patterns on PCBs) that realize these impedance distributions.

Current Challenges:

Computational Cost: Traditional microscopic design relies on time-consuming full-wave electromagnetic (EM) optimization (e.g., using CST or HFSS) for every unit cell, which is inefficient for large arrays.
Accuracy Limitations: Existing analytical models (like Transmission Line Models) often neglect near-field interlayer coupling, leading to significant inaccuracies, especially in ultracompact designs where dielectric layers are thin relative to the meta-atom size.
Dual-Polarization & Bandwidth: Creating designs that work for both TE and TM polarizations and maintain performance over a wide frequency band is difficult with standard analytical approaches.
Fabrication Readiness: Many theoretical models do not account for realistic PCB constraints (dielectric stacks, conductor losses), making the transition from theory to physical fabrication challenging.

2. Methodology

The authors propose a semianalytical framework codified in a MATLAB program called LAYERS. This approach bridges the gap between rigorous analytical speed and full-wave accuracy.

A. Rigorous Analytical Model (LAYERS)

Extension of Metagrating Theory: The model extends previous analytical formulations for sparse loaded-wire metagratings to densely packed Jerusalem Cross (JC) meta-atoms embedded in standard multilayer PCB dielectric stacks.
Floquet-Bloch Analysis: It solves for scattering from periodic multiple layers of dielectric substrates with etched metallic patterns. It accounts for:
- Both propagating and evanescent modes (capturing near-field interlayer coupling).
- Realistic conductor and dielectric losses.
- Mutual interactions between scatterers in all layers.
Dual-Polarization: By utilizing JC patterns (which possess 90° rotational symmetry), the model handles both TE and TM polarizations simultaneously.
Effective Load Impedance: Instead of modeling complex JC geometries directly in the analytical solver, the authors derive a functional relationship $Z_n = Z_n(W_n)$ $Z_{n} = Z_{n} (W_{n})$ , where $Z_n$ $Z_{n}$ is the distributed impedance and $W_n$ $W_{n}$ is the leg length of the JC. This is achieved by:
1. Running a limited number of full-wave simulations (CST) for a single interface to map JC leg lengths to effective impedances.
2. Fitting these results to polynomial functions.
3. Feeding these functions into the LAYERS solver to predict the transmission coefficient $T$ for any combination of leg lengths across the stack.

B. Lookup Table (LUT) Construction

Exhaustive Sampling: The authors perform an exhaustive search over the leg-length parameters ( $W_n$ ) of the JC patterns.
Filtering: They identify configurations that yield high transmission magnitude ( $|T| \approx 1$ ) across the full $2\pi$ phase range.
Hybrid Validation: While LAYERS provides the initial candidates, the final LUT entries are verified using full-wave solvers (CST) to correct for minor discrepancies caused by higher-order multipole effects neglected in the dipole approximation. This step is computationally efficient because the search space is pre-filtered by the analytical model.

C. Frequency Response Scaling

A judicious scaling method is introduced to predict wideband behavior without re-running full-wave simulations for every frequency.
The model treats structural lengths as relative to a unit wavelength. To analyze a new frequency, the physical dimensions are scaled inversely to the wavelength change, allowing rapid prediction of meta-atom performance across a frequency band.

3. Key Contributions

Semianalytical Synthesis Scheme: A novel method that eliminates the need for full-wave optimization in the microscopic design stage, drastically reducing design time while maintaining high accuracy.
Dual-Polarized PCB Compatibility: The framework specifically targets fabrication-ready multilayer PCBs using Jerusalem Cross meta-atoms, enabling dual-polarized operation (TE and TM) with low reflection.
Wideband Prediction: The introduction of a scaling law allows the analytical model to predict frequency responses over a broad bandwidth (demonstrated over a 10% fractional bandwidth).
Optimization of Layer Thickness: The study demonstrates that layer thicknesses can be treated as design degrees of freedom to optimize mean transmission efficiency across a frequency band, improving performance by over 15% compared to standard stack configurations.
Open-Source Toolkit: The authors provide the LAYERS code and the resulting LUTs, establishing a foundation for the companion paper (Part II) which integrates machine learning (MetaMamba) for even faster inverse design.

4. Results

Validation: The analytical predictions from LAYERS showed excellent agreement with full-wave CST simulations for both transmission magnitude and phase.
Case Study (Metalens):
- A dual-polarized Huygens' metalens was designed to convert a normally incident plane wave into a cylindrical wave (focusing at $y_c = 3\lambda$ ).
- Performance:
  - TM Polarization: ~80.0% of incident power reached the focal plane; ~5.8% reflected; ~14.2% lost.
  - TE Polarization: ~83.2% of incident power reached the focal plane; ~4.8% reflected; ~12.0% lost.
  - Focus Quality: The Full Width at Half Maximum (FWHM) was approximately $0.46\lambda$ for both polarizations, indicating a sharp, well-formed focus.
Wideband Efficiency: By optimizing the laminate thickness ( $h_2, h_3$ ), the average mean transmission efficiency over a 18–22 GHz band was significantly improved, demonstrating the method's utility for broadband applications.

5. Significance

Bridging Theory and Practice: This work solves the "accuracy vs. speed" trade-off in metasurface design. It provides a rigorous analytical tool that accounts for realistic PCB constraints (losses, coupling) without the computational burden of full-wave optimization.
Enabling Rapid Prototyping: The ability to generate a fabrication-ready LUT in minutes (on a standard PC) rather than days accelerates the development of complex wave-manipulating devices like metalenses, beam steerers, and holograms.
Foundation for AI: Part I establishes the high-fidelity "ground truth" data generation engine. This is critical for Part II, where a Selective State Space Model (Mamba) is trained on this semianalytical data and fine-tuned with minimal full-wave runs to achieve ultrafast inverse design.
Open Science: By releasing the code and methodology, the authors democratize access to advanced metasurface synthesis tools, fostering further innovation in the field of metamaterials and microwave engineering.

In summary, this paper presents a robust, semianalytical methodology that enables the rapid, accurate, and fabrication-ready design of dual-polarized Huygens' metasurfaces, overcoming the limitations of traditional full-wave optimization and simplified analytical models.