CS Paper Radar

Diffusion-Based Inverse Design of Dielectric Resonator Metasurfaces for Shaping Smart Electromagnetic Environments

M. Tsukerman, K. Grotov, D. Vovchuk, P. Ginzburg

cs.LG physics.app-phphysics.comp-ph ·submitted 2026-08-30 17:04 UTC · v1

Abstract

Future wireless systems are expected to transform the surrounding space from a passive propagation medium into a smart electromagnetic environment, where engineered surfaces control wave propagation, support wireless sensing, and create programmable electromagnetic fingerprints. A key challenge in realizing this vision is the inverse design of metasurfaces for tailored electromagnetic propagation. While forward analysis evaluates the response of a known geometry, the inverse task starts from a prescribed scattering signature and seeks a physically realizable structure that produces it. This inverse task is inherently nonlinear and often high-dimensional, while candidate solutions may be non-unique and provide no direct indication of practical realizability. Here, we introduce a conditional diffusion framework for inverse design of dielectric resonator metasurfaces from target angular scattering patterns. Trained on T-matrix simulated geometry-response pairs, the model learns a conditional distribution of geometries instead of a deterministic mapping, enabling multiple candidate designs for the ill-posed inverse problem. The best generated metasurface achieves a mean percentage error of 1.39%, outperforming CMA-ES optimization (4.1% after 10 h) while requiring only about one minute for after-training inference. The model also produces lower error distributions than deterministic neural baselines for out-of-distribution spectra, highlighting the potential of diffusion models for efficient metasurface design.

Closest papers

Nearest neighbours by meaning across the whole index.

Multi-Solution Inverse Design of Pixelated Patch Antennas Using an Autoregressive Transformer

cs.LG ·2026-10-29 ·Young-Mok Yun, Seung-Won Shin, Jaeyong Chung, Kyung-Sik Choi
similarity 0.59
Abstract · 16 pages, 28 figures, 7 tables

Deep-learning-based inverse design has emerged as a promising framework for pixelated patch antenna synthesis by enabling broader exploration of discrete geometry spaces and reducing design time compared with conventional template-based methods. Existing methods, including surrogate-assisted metaheuristic optimization, tandem inverse networks, and diffusion models, have advanced the field but still face practical limitations. Metaheuristic methods such as binary particle swarm optimization (BPSO) are sensitive to initialization and search hyperparameters and are often time-consuming, tandem networks have difficulty capturing the one-to-many nature of inverse antenna design, and diffusion models may require excessive complexity and data for moderate-scale binary layout synthesis. In this paper, an autoregressive (AR) transformer is presented for the inverse design of 10 x 10 pixelated patch antennas. Its sequential discrete generation process enables the synthesis of multiple feasible structures for a given target response. The pretrained model is further fine-tuned for dual-band design using step-shaped target responses with sharp spectral transitions, allowing intuitive band specification without manually constructing detailed S11 curves. Compared with BPSO, our proposed method generates valid and diverse antenna candidates more efficiently. Fabricated and measured examples confirm that distinct pixelated layouts can realize the desired antenna responses, thereby demonstrating the multi-solution capability of the proposed framework.

0.25

Möbius Reparametrization of Multiport-Network Models of PIN-Diode-Programmable Metasurfaces for Accurate Low-Order Neumann Approximations

eess.SP physics.app-ph ·2026-09-29 ·Philipp del Hougne
similarity 0.59
Abstract · 6 pages with 3 figures

Accurate models of programmable metasurfaces based on multiport-network theory (MNT) account for mutual coupling (MC) through a configuration-dependent matrix inversion. The latter's high computational cost during gradient-based optimization (GBO) can be alleviated via a finite-order Neumann approximation. We show that the accuracy of this approximation depends strongly on the (tacitly) chosen MNT parametrization. For 1-bit-programmable meta-elements (e.g., meta-elements based on PIN diodes), we identify a closed-form Möbius reparametrization that depends only on the meta-element's two load states and requires neither training data nor numerical optimization. For an experimentally estimated proxy MNT model of a fabricated 96-element 19-GHz dynamic metasurface antenna with strong MC, a second-order Neumann approximation with our reparametrization achieves forward and control-gradient accuracies of 21.68 and 21.70 dB, respectively. Relative to the full proxy MNT, it reduces the runtime of a combined forward and control-gradient evaluation by a third and the saved-tensor memory by a factor of ten. In a prototypical GBO problem of end-to-end optimization for DMA-based scene classification, it achieves 94.97% test accuracy vs. 95.57% with the full model. We further note that MC-strength metrics and MC-unaware benchmarks should be parametrization-invariant, motivating, for instance, definitions based on the best directly fitted zeroth-order model.

0.48

Simulator-Refined Diffusion for Radio-Frequency Inverse Design

cs.LG cs.AI ·2026-09-29 ·Jinhao Liang, Jacob K. Christopher, Michael Frei, Tommaso Dreossi, Nando Fioretto
similarity 0.58
Abstract

Diffusion models have shown potential in inverse design of printed circuit boards (PCBs), enabling the generation of layouts conditioned on target S-parameters. Despite this promise, applying diffusion models to PCB layout generation remains challenging due to their difficulty in meeting the quantitative electromagnetic specifications. A common approach is gradient-based guidance, which biases the diffusion sampling process with the gradient of an objective used for evaluation. However, full-wave electromagnetic simulators are accurate but expensive and typically non-differentiable, whereas differentiable surrogates are informative but not always reliable. To address these limitations, this paper proposes Simulator-Refined Diffusion (SRD), a novel combination of a low-fidelity differentiable surrogate and a high-fidelity non-differentiable simulator within the diffusion sampling process. Unlike standard zeroth-order optimization, which requires a great number of random perturbations, our approach uses the surrogate's gradient to propose the perturbation direction while the simulator then searches based on this direction to identify an effective design update. Experimental results across different settings show that this method consistently outperforms current state-of-the-art methods, producing layouts whose simulated S-parameters match the target specifications up to 21.2% closer for in-distribution targets and up to 19.8% for out-of-distribution targets.

0.15

Electromagnetic Information Theory for Programmable Metasurfaces

eess.SP physics.optics ·2026-11-02 ·Philipp del Hougne, Marco Di Renzo, Owen D. Miller, David A. B. Miller
position similarity 0.58
Abstract · 17 pages with 3 figures

Programmable metasurfaces enable fine-grained control of electromagnetic waves and are creating new opportunities for information extraction, information transfer and information processing. Yet, despite exciting experimental demonstrations, their fundamental capabilities remain poorly understood. In this Perspective, we argue that programmable metasurfaces, and programmable wave systems more generally, call for an electromagnetic information theory that addresses limits at two levels, in a unified manner across applications like communications, sensing and computing. First, for a given physical platform, such a theory should bound achievable information manipulations, such as sensing precision, information-transfer rates, and computational performance. Second, at a more fundamental level, it should determine what non-programmable and programmable physical resources bound these system-specific limits, and what minimum resources are required to attain a prescribed level of performance. We discuss emerging theoretical approaches to these questions, and their connection to experimentally realizable systems. We examine how device thickness, programmable degrees of freedom, nonlocality, coupling and multiple scattering constrain these limits. Establishing the envisioned framework could clarify the ultimate capabilities of programmable metasurfaces and guide the design of future wave-based information systems.

0.27

Goal-Oriented Wave Computing for Direction-of-Arrival Estimation

eess.SP physics.app-ph ·2026-10-05 ·Mert Kalfa
benchmark similarity 0.57
Abstract · 9 pages, 2 figures, 4 tables

Programmable metasurfaces---arrays of elements whose electromagnetic response can be tuned electronically---can process signals as a wave propagates through or reflects off them, a paradigm known as wave-domain analog computing. The design strategy is typically indirect: the surface is fitted to a prescribed linear operator, such as a discrete Fourier transform, and the result is read from output power measurements. Using direction-of-arrival (DoA) estimation as a canonical benchmark, we study what this indirection costs across analog computing topologies, namely reflective reconfigurable intelligent surfaces (RIS), beyond-diagonal RIS (BD-RIS), and stacked intelligent metasurfaces (SIM). We first derive Cramér--Rao bounds for the complete physical estimation chain, including one that accounts for output detectors measuring only intensity. We then formulate a goal-oriented alternative where the surfaces are trained directly through the deployed readout, with no operator target. We observe that goal-oriented training lets sparsely connected or shallower surfaces reach the task performance of their more complex operator-fitted counterparts---despite nearly identical Fisher information---and that the advantage persists under coarsely quantized (1--3 bit) control. From 16 to 64 elements, the depth saving persists, while the connectivity a beyond-diagonal RIS needs grows with the aperture. Under tuner and per-layer loss, the goal-oriented designs keep their advantage.

0.25

Physics-Augmented Graph Transformers for Patch-Antenna Forward and Inverse Design

cs.LG physics.comp-ph ·2026-10-04 ·Avi Epstein, Snir Nehemia, Haim Suchowski, Lior Wolf
publishedbenchmarkcode/data similarity 0.56
Abstract · 6 pages, 3 figures, 3 tables. Accepted for oral presentation at the 2026 IEEE International Workshop on Machine Learning for Signal Processi…

Full-wave electromagnetic (EM) simulation enables accurate patch-antenna analysis but is computationally expensive for large-scale forward prediction and inverse design. We present a mesh-native, physics-augmented graph-learning framework that treats radiation-pattern prediction as signal reconstruction on an irregular surface mesh. For the forward problem, a GPS graph transformer is trained with Physics-Augmented Intermediate Supervision (PAIS), an auxiliary node-level objective that predicts complex surface currents, the physical intermediate linking geometry to radiation. PAIS improves multiple GNN backbones at no inference-time cost, while shuffled-current and non-physical controls show the gain comes from physical correspondence. Direction-conditioned decoding and a differentiable radiation-integral consistency loss further exploit this structure. On an 80,000-sample CST benchmark, GPS+PAIS reaches MSE 0.17 / PSNR 19.67, generalizes to a PCA split, and transfers zero-shot to canonical patches. For inverse design, surrogate-filtered diffusion beats nearest-neighbor retrieval by 32% relative MSE.

0.47

Inductively Scalable, Single-Step Neural Surrogates for Wave-Scattering Inverse Problems

cs.AI physics.optics ·2026-08-18 ·Charles Dove, Laura Waller
similarity 0.56
Abstract

Neural network surrogates are an emerging alternative to traditional electromagnetic wave simulators like finite-difference time-domain (FDTD); their goal is to replace rigorous physical simulations with pre-trained neural networks that solve wave-scattering forward and inverse problems orders of magnitude faster. However, nonrecurrent, single-step surrogates have scaled only to a few tens of simulation variables. Here, we show that this barrier can be overcome by dynamically generating salient training examples during training, rather than randomly sampling the large space of possible examples. We introduce an algorithm that runs in parallel with surrogate training, using gradient ascent to search refractive-index and source configurations for cases where the surrogate disagrees with a full-wave ground-truth simulator. We also use source and ground-truth normalization with an evolving replay dataset to stabilize and accelerate learning. Using this approach, we train a fast, single-step surrogate for two-dimensional wave scattering with up to 41,772 controllable variables, including dense, freely configurable grids of refractive indices and complex-valued sources. The resulting neural surrogate is robustly accurate across diverse structured and unstructured examples and generalizes inductively to larger domains, reaching over 3 million controllable variables without retraining, a $73.8\times$ increase. We demonstrate the surrogate on large-scale forward simulations and inverse design of freeform beam splitters and gradient-index (GRIN) lenses up to 98 wavelengths wide, showing comparable or better performance than FDTD-based designs, with speedups from $1.29\times$ to $26.5\times$. These results demonstrate a practical path toward fast, robustly accurate, inductively scalable neural simulators for photonic inverse design and other wave-scattering inverse problems.

0.24

Dynamic Metasurface Antennas: From Programmable Microwave Hardware to Next-Generation Wireless Systems

eess.SP ·2026-09-24 ·Abdul Jabbar, William Whittow
similarity 0.56
Abstract

As wireless systems evolve toward 6G and beyond, antenna apertures are increasingly expected to become more programmable, multifunctional, and responsive to diverse communication, sensing, and imaging requirements. Dynamic metasurface antennas (DMAs) offer a promising hardware pathway toward this vision by integrating electronically addressable metamaterial elements within microwave and millimeter-wave (mmWave) radiating apertures. Unlike conventional phased arrays relying on dedicated phase-shifting networks, DMAs manipulate guided waves in microwave structures through electronically controlled resonant elements. In this way, DMAs offer phase-shifter-less beam-steering, multibeam radiation, and programmable wavefront control. This article explores state-of-the-art DMA technology from a microwave-engineering perspective, spanning guided-wave excitation, reconfigurable resonant meta-elements, microwave tuning technologies, RF-digital integration, and FPGA-controlled beamforming. Closely related electronically steered metasurface antenna architectures are also discussed within a unified microwave-hardware framework. Recent advances illustrating how guided-wave DMA hardware and high-speed electronics facilitate programmable communication, sensing, imaging, and space-time functionalities are reviewed. Key microwave design challenges toward wideband, scalable, and highly integrated DMA apertures are highlighted. Finally, emerging opportunities in satellite communications, near-field connectivity, fixed wireless access, wave-domain computing, and integrated wireless functionalities are explored as future directions for DMA technology. This article provides a unified microwave engineering perspective on the design, practical implementation, and emerging capabilities of DMAs as programmable front-ends for next-generation wireless systems.

0.26