Pith. sign in

REVIEW 1 major objections 1 minor 46 references

ADC-Aware End-to-End Optimization of a Dynamic Metasurface Antenna with Strong Mutual Coupling for Monostatic Scene Classification

T0 review · 1 major / 1 minor · reviewed 2026-07-02 · grok-4.3

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read ADC-aware end-to-end training of a mutual-coupling DMA model preserves 87% accuracy under one-bit quantization while ideal-ADC training collapses to 56%.

desk verdict ADC-aware training recovers most accuracy lost to 1-bit quantization in this DMA setup, but the gains sit inside a simulation whose fitted multiport model lacks reported cross-checks against the actual device on the task data. read the letter →

arxiv 2607.00253 v1 pith:QP7HCATE submitted 2026-06-30 eess.SP physics.app-ph

Philipp del Hougne This is my paper
classification eess.SPphysics.app-ph
keywords dynamicmetasurfaceantennamutualcouplingADC-awareoptimizationmonostaticsensingsceneclassificationone-bitADCend-to-endmultiportnetworkmodel
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper demonstrates that jointly optimizing the configurations of a 96-element dynamic metasurface antenna and a downstream digital classifier, while explicitly modeling both strong mutual coupling and the quantization effects of low-resolution ADCs, enables reliable monostatic scene classification. Using an experimentally calibrated multiport network model, the approach achieves 87.2% test accuracy with eight DMA states and uniform one-bit ADCs. In contrast, a system trained under ideal-ADC assumptions drops to 56% when the same uniform one-bit ADC is applied at test time, and omitting the mutual-coupling model reduces performance to random-guess levels. Learning non-uniform ADC thresholds yields only modest further gains.

What carries the argument

MC-aware multiport-network model with experimentally estimated parameters that maps the 96 one-bit-programmable meta-element states to the wave-domain signals before ADC quantization.

What would settle it

A direct measurement of the fabricated DMA's radiated fields for the eight trained configurations, under the same scene and frequency conditions, that deviates significantly from the multiport-network model predictions would falsify the reported accuracy numbers.

Watch

Extended reading notes

Core claim

ADC-aware end-to-end optimization of DMA configurations and the digital classifier, performed with an MC-aware multiport-network model whose parameters were experimentally estimated on a fabricated chaotic-cavity-backed DMA, maintains high classification accuracy even when one-bit ADCs are present; the same pipeline trained under an ideal-ADC assumption suffers severe degradation, and ignoring mutual coupling reduces accuracy to chance.

Load-bearing premise

The experimentally estimated parameters of the MC-aware multiport-network model accurately represent the fabricated DMA's electromagnetic response under the exact operating conditions and scene-classification task used for training and testing.

Editorial extensions

If this is right

  • ADC awareness during training prevents accuracy collapse from 95.5% to 56% when one-bit ADCs are deployed.
  • Mutual-coupling modeling is indispensable; its absence reduces performance to random guessing.
  • Jointly learning non-uniform ADC thresholds provides at most modest gains over fixed uniform thresholds.
  • End-to-end optimization of DMA states and classifier remains effective for monostatic sensing pipelines once hardware non-idealities are included in the model.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Similar hardware-aware modeling could improve performance in DMA-based communication or imaging tasks that also face low-resolution ADCs.
  • The approach suggests that any programmable metasurface sensing system should incorporate measured mutual-coupling matrices rather than idealized isolated-element assumptions.
  • Extending the framework to jointly optimize over a larger set of DMA configurations or continuous phase states may further close the gap to ideal-ADC performance.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

1 major / 1 minor

Summary. The manuscript studies ADC-aware end-to-end optimization of a monostatic scene-classification pipeline that uses a dynamic metasurface antenna (DMA) with strong mutual coupling. An MC-aware multiport-network model is constructed from experimentally estimated parameters of a fabricated 96-element chaotic-cavity-backed DMA. End-to-end training jointly optimizes DMA configurations and a digital classifier, either with a fixed uniform ADC or with learned non-uniform thresholds, and is compared to ideal-ADC and MC-ignorant baselines. The central numerical result is that, for one-bit ADCs and eight DMA configurations, an ideal-ADC-trained system drops from 95.5% to 56.0% test accuracy when a uniform one-bit ADC is applied at inference, while ADC-aware training recovers 87.2% accuracy; ignoring mutual coupling reduces performance to chance level.

Significance. If the calibrated model is faithful under the exact operating conditions of the task, the work supplies clear, quantitative evidence that hardware non-idealities (ADC quantization and mutual coupling) must be included in the optimization loop for DMA-based sensing. The use of measured parameters rather than purely analytic models is a strength, and the explicit baseline comparisons make the accuracy deltas directly interpretable.

major comments (1)
  1. [model calibration and results sections (accuracy comparisons)] The headline accuracy deltas (95.5 % o 56.0 % vs. 87.2 %) are generated entirely inside a simulation driven by a single set of fitted multiport parameters. The manuscript does not report any cross-validation of the model’s predicted radiated fields or coupling coefficients against fresh measurements taken on the same DMA configurations, frequencies, and scene geometries that appear in the training and test sets. Because any unmodeled frequency dependence, nonlinearity, or fabrication drift would directly alter the reported benefit of ADC-aware training, this verification step is load-bearing for the claim that ADC awareness is essential in a real monostatic sensing pipeline.
minor comments (1)
  1. [experimental results] Dataset size, number of scenes per class, and any statistical significance testing on the reported accuracy figures are not mentioned in the abstract and should be stated explicitly in the experimental section.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback and for recognizing the value of including hardware non-idealities in the optimization. We address the single major comment below.

read point-by-point responses

  1. Referee: [model calibration and results sections (accuracy comparisons)] The headline accuracy deltas (95.5 % o 56.0 % vs. 87.2 %) are generated entirely inside a simulation driven by a single set of fitted multiport parameters. The manuscript does not report any cross-validation of the model’s predicted radiated fields or coupling coefficients against fresh measurements taken on the same DMA configurations, frequencies, and scene geometries that appear in the training and test sets. Because any unmodeled frequency dependence, nonlinearity, or fabrication drift would directly alter the reported benefit of ADC-aware training, this verification step is load-bearing for the claim that ADC awareness is essential in a real monostatic sensing pipeline.

    Authors: We agree that additional cross-validation of the model predictions against fresh measurements on the exact configurations, frequencies, and scene geometries would strengthen the results. The multiport parameters were obtained via experimental calibration on the fabricated DMA, as stated in the manuscript; the present work uses this calibrated model to quantify the impact of ADC quantization and mutual coupling within a realistic simulation. We will revise the manuscript to add an explicit discussion of the model assumptions, the scope of the experimental calibration, and the fact that full end-to-end experimental verification lies beyond the current study. revision: yes

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity; accuracies are simulation outcomes from experimentally fitted model, not algebraic reductions

full rationale

The paper reports accuracy deltas (95.5% ideal-ADC-trained vs. 56.0% and 87.2% for one-bit ADC cases) as results of end-to-end optimization performed inside a multiport-network model whose parameters were obtained via separate experimental estimation on the fabricated DMA. These numbers are not forced by construction from the fitted parameters; they emerge from comparing distinct training regimes (ideal-ADC assumption vs. ADC-aware) and baselines (with/without MC awareness). No self-definitional equations, fitted-input-as-prediction steps, or load-bearing self-citations appear in the abstract or described chain. The derivation is therefore self-contained against external measurement benchmarks and receives the default low score.

Assumptions & free parameters 1 free parameters · 1 assumptions · 0 invented entities

Central claim rests on the fidelity of the experimentally estimated multiport parameters and on the assumption that the chosen scene-classification task and eight-configuration budget are representative of real deployment.

free parameters (1)
  • multiport network parameters
    Experimentally estimated from measurements on the fabricated 96-element DMA
assumptions (1)
  • domain assumption The multiport-network model with measured parameters faithfully reproduces the DMA's wave-domain behavior for the monostatic sensing task
    Invoked to generate the training data for end-to-end optimization

how reviews work

0 comments
Cite this review

Pith. "Pith review of ADC-Aware End-to-End Optimization of a Dynamic Metasurface Antenna with Strong Mutual Coupling for Monostatic Scene Classification." pith.science (2026). https://pith.science/paper/QP7HCATE

@misc{pith2026260700253,
  author       = {Pith},
  title        = {Pith review of: ADC-Aware End-to-End Optimization of a Dynamic Metasurface Antenna with Strong Mutual Coupling for Monostatic Scene Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QP7HCATE}},
  note         = {Machine review of arXiv:2607.00253}
}
read the original abstract

Dynamic metasurface antennas (DMAs) enable programmable wave-domain signal processing that can be jointly optimized with downstream digital processing in an end-to-end manner. Existing studies, however, typically assume ideal analog-to-digital conversion (ADC) and often rely on simplified electromagnetic models. Here, we study ADC-aware end-to-end optimization of a monostatic sensing pipeline based on a DMA with strong mutual coupling (MC). We model the wave domain using an MC-aware multiport-network model whose parameters were experimentally estimated for a fabricated chaotic-cavity-backed DMA with 96 one-bit-programmable meta-elements. We perform ADC-aware end-to-end optimization of the DMA configurations and digital classifier, either with awareness of a fixed uniform ADC or, optionally, with jointly learned ADC decision thresholds, and compare against baselines that assume an ideal ADC and/or ignore MC. Our results show that ADC awareness is essential in low-resolution ADC regimes: with one-bit ADCs and eight DMA configurations, deploying an ideal-ADC-trained system with a uniform one-bit ADC reduces the test accuracy from 95.5% to 56.0%, whereas ADC-aware training with the same fixed uniform one-bit ADC achieves 87.2%. We also show that without MC awareness the accuracy drops to the random-guess level. Learning non-uniform ADC thresholds provides at most modest additional gains over fixed uniform ADCs in the considered DMA-based sensing pipeline.

Figures

Figures reproduced from arXiv: 2607.00253 by the authors.

Figure 1
Figure 1. Schematic overview of the considered end-to-end scene-classification sensing pipeline, comprising the analog wave domain, the analog-to-digital conversion, [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Schematic illustration of the MNT-based model for a generic multi-feed [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Soft-to-hard DMA relaxation for a representative [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Illustration of hard ADC quantization and its differentiable relaxation for [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: DMA prototype and VAA-based experimental characterization setup. [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Scene-classification accuracy with one-bit ADCs as a function of the number of DMA configurations [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: Test-set ADC inputs for M = 1, shown in the normalized I-Q plane and colored by MNIST class. Rows compare random and learned DMA configurations; columns compare ideal, uniform 2-bit, and learned 2-bit ADCs. For finite-resolution ADCs, solid black lines mark internal bi…
Figure 8
Figure 8. Figure 8: PDFs of the normalized I and Q ADC inputs corresponding to the test-set clouds shown in Fig. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]
Figure 9
Figure 9. Figure 9: Scene-classification accuracy for the single-shot case [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

46 extracted references · 46 canonical work pages

  1. [1]

    Programmable wave-domain computing in wireless communications,

    P. del Hougneet al., “Programmable wave-domain computing in wireless communications,”HAL:05487878, 2026

    work page 2026

  2. [2]

    Dynamic metamaterial aperture for microwave imaging,

    T. Sleasmanet al., “Dynamic metamaterial aperture for microwave imaging,” Appl. Phys. Lett., vol. 107, no. 20, 2015

    work page 2015

  3. [3]

    Implementation and characterization of a two- dimensional printed circuit dynamic metasurface aperture for computational microwave imaging,

    T. A. Sleasmanet al., “Implementation and characterization of a two- dimensional printed circuit dynamic metasurface aperture for computational microwave imaging,”IEEE Trans. Antennas Propag., vol. 69, no. 4, pp. 2151–2164, 2020

    work page 2020

  4. [4]

    Beyond-diagonal dynamic metasurface antenna,

    H. Prod’homme and P. del Hougne, “Beyond-diagonal dynamic metasurface antenna,”IEEE Commun. Lett., vol. 30, pp. 258–262, 2025

    work page 2025

  5. [5]

    RF chain reduction for MIMO systems: A hardware prototype,

    T. Gonget al., “RF chain reduction for MIMO systems: A hardware prototype,”IEEE Sys. J., vol. 14, no. 4, pp. 5296–5307, 2020

    work page 2020

  6. [6]

    Dynamic metasurface antennas for 6G extreme massive MIMO communications,

    N. Shlezingeret al., “Dynamic metasurface antennas for 6G extreme massive MIMO communications,”IEEE Wirel. Commun., vol. 28, no. 2, pp. 106–113, 2021

    work page 2021

  7. [7]

    Learned integrated sensing pipeline: Reconfigurable metasurface transceivers as trainable physical layer in an artificial neural network,

    P. del Hougneet al., “Learned integrated sensing pipeline: Reconfigurable metasurface transceivers as trainable physical layer in an artificial neural network,”Adv. Sci., vol. 7, no. 3, p. 1901913, 2020

    work page 2020

  8. [8]

    Noise-adaptive intelligent programmable meta-imager,

    C. Qian and P. del Hougne, “Noise-adaptive intelligent programmable meta-imager,”Intell. Comput., 2022

    work page 2022

Show all 46 references
  1. [9]

    On low-resolution ADCs in practical 5G millimeter- wave massive MIMO systems,

    J. Zhanget al., “On low-resolution ADCs in practical 5G millimeter- wave massive MIMO systems,”IEEE Commun. Mag., vol. 56, no. 7, pp. 205–211, 2018

    2018

  2. [10]

    Analog-to-digital compression: A new paradigm for converting signals to bits,

    A. Kipniset al., “Analog-to-digital compression: A new paradigm for converting signals to bits,”IEEE Signal Process. Mag., vol. 35, no. 3, pp. 16–39, May 2018

    2018

  3. [11]

    Hardware-limited task-based quantization,

    N. Shlezingeret al., “Hardware-limited task-based quantization,”IEEE Trans. Signal Process., vol. 67, no. 20, pp. 5223–5238, Oct. 2019

    2019

  4. [12]

    Asymptotic task-based quantization with application to massive MIMO,

    N. Shlezingeret al., “Asymptotic task-based quantization with application to massive MIMO,”IEEE Trans. Signal Process., vol. 67, no. 15, pp. 3995–4012, Aug. 2019

    2019

  5. [13]

    Task-based analog-to-digital converters,

    P. Neuhauset al., “Task-based analog-to-digital converters,”IEEE Trans. Signal Process., vol. 69, pp. 5403–5418, 2021

    2021

  6. [14]

    Dynamic metasurface antennas for MIMO-OFDM receivers with bit-limited ADCs,

    H. Wanget al., “Dynamic metasurface antennas for MIMO-OFDM receivers with bit-limited ADCs,”IEEE Trans. Commun., vol. 69, no. 4, pp. 2643– 2659, Apr. 2021

    2021

  7. [15]

    Dynamic metasurface antenna based anti-jamming with bit-limited ADC,

    Y . Haoet al., “Dynamic metasurface antenna based anti-jamming with bit-limited ADC,”IEEE Trans. Veh. Technol., vol. 75, no. 1, p. 1603, 2026

    2026

  8. [16]

    Deep task-based analog-to-digital conversion,

    N. Shlezingeret al., “Deep task-based analog-to-digital conversion,”IEEE Trans. Signal Process., vol. 70, pp. 6021–6034, 2022

    2022

  9. [17]

    Learning task-based trainable neuromorphic ADCs via power- aware distillation,

    T. V olet al., “Learning task-based trainable neuromorphic ADCs via power- aware distillation,”IEEE Trans. Signal Process., vol. 73, pp. 1246–1261, 2025

    2025

  10. [18]

    Mutual coupling in dynamic metasurface antennas: Foe, but also friend,

    H. Prod’homme and P. del Hougne, “Mutual coupling in dynamic metasurface antennas: Foe, but also friend,”IEEE Wirel. Commun., vol. 32, no. 4, pp. 30–36, 2025

    2025

  11. [19]

    Benefits of mutual coupling in dynamic metasurface antennas,

    H. Prod’hommeet al., “Benefits of mutual coupling in dynamic metasurface antennas,”IEEE Trans. Antennas Propag., vol. 74, no. 3, p. 2589, 2025

    2025

  12. [20]

    Electromagnetic based communication model for dynamic metasurface antennas,

    R. J. Williamset al., “Electromagnetic based communication model for dynamic metasurface antennas,”IEEE Trans. Wirel. Commun., vol. 21, no. 10, pp. 8616–8630, 2022

    2022

  13. [21]

    Network-based design of reactive beamforming metasurfaces,

    M. Almunifet al., “Network-based design of reactive beamforming metasurfaces,”IEEE Trans. Antennas Propag., vol. 73, p. 5970, 2025

    2025

  14. [22]

    Metasurface-based fluid antennas: from electromagnetics to communications model,

    P. Ramírez-Espinosaet al., “Metasurface-based fluid antennas: from electromagnetics to communications model,”arXiv:2507.17982, 2025

    2025

  15. [23]

    Experimental multiport-network parameter estimation for a dynamic metasurface antenna,

    J. Tapie and P. del Hougne, “Experimental multiport-network parameter estimation for a dynamic metasurface antenna,”IEEE Trans. Antennas Propag., 2026

    2026

  16. [24]

    Intelligent meta-imagers: From compressed to learned sensing,

    C. Saigre-Tardifet al., “Intelligent meta-imagers: From compressed to learned sensing,”Appl. Phys. Rev., vol. 9, no. 1, 2022

    2022

  17. [25]

    Transmission-type 2-bit programmable metasurface for single-sensor and single-frequency microwave imaging,

    Y . B. Liet al., “Transmission-type 2-bit programmable metasurface for single-sensor and single-frequency microwave imaging,”Sci. Rep., vol. 6, no. 1, p. 23731, 2016

    2016

  18. [26]

    Intelligent electromagnetic sensing with learnable data acquisition and processing,

    H.-Y . Liet al., “Intelligent electromagnetic sensing with learnable data acquisition and processing,”Patterns, vol. 1, no. 1, 2020

    2020

  19. [27]

    Sensing matrix design via mutual coherence minimization for electromagnetic compressive imaging applications,

    R. Obermeier and J. A. Martinez-Lorenzo, “Sensing matrix design via mutual coherence minimization for electromagnetic compressive imaging applications,”IEEE Trans. Comput. Imaging, vol. 3, no. 2, p. 217, 2017

    2017

  20. [28]

    Sensing matrix design via capacity maximization for block compressive sensing applications,

    ——, “Sensing matrix design via capacity maximization for block compressive sensing applications,”IEEE Trans. Comput. Imaging, vol. 5, no. 1, pp. 27–36, 2019

    2019

  21. [29]

    Digitized metamaterial absorber-based compressive reflector antenna for high sensing capacity imaging,

    A. Molaeiet al., “Digitized metamaterial absorber-based compressive reflector antenna for high sensing capacity imaging,”IEEE Access, vol. 7, pp. 1160–1173, 2019

    2019

  22. [30]

    Optimal multiplexing of spatially encoded information across custom-tailored configurations of a metasurface-tunable chaotic cavity,

    P. del Hougneet al., “Optimal multiplexing of spatially encoded information across custom-tailored configurations of a metasurface-tunable chaotic cavity,”Phys. Rev. Appl., vol. 13, no. 4, p. 041004, 2020

    2020

  23. [31]

    Generalized optimization of high-capacity compressive imaging systems,

    R. Obermeier and J. A. Martinez-Lorenzo, “Generalized optimization of high-capacity compressive imaging systems,”IEEE Trans. Antennas Propag., vol. 68, no. 4, pp. 3147–3161, 2020

    2020

  24. [32]

    Optimizing dynamic metasurface antenna configurations for direction-of-arrival and polarization estimation using an experimentally calibrated multiport-network model,

    J. Tapie and P. del Hougne, “Optimizing dynamic metasurface antenna configurations for direction-of-arrival and polarization estimation using an experimentally calibrated multiport-network model,”Opt. Mater. Exp., vol. 16, no. 7, pp. 2016–2031, 2026

    2016

  25. [33]

    Reconfigurable array design to realize principal component analysis (PCA)-based microwave compressive sensing imaging system,

    M. Lianget al., “Reconfigurable array design to realize principal component analysis (PCA)-based microwave compressive sensing imaging system,” IEEE Antennas Wirel. Propag. Lett., vol. 14, pp. 1039–1042, 2015

    2015

  26. [34]

    Machine-learning reprogrammable metasurface imager,

    L. Liet al., “Machine-learning reprogrammable metasurface imager,”Nat. Commun., vol. 10, no. 1, p. 1082, 2019

    2019

  27. [35]

    Learning sensor multiplexing design through back- propagation,

    A. Chakrabarti, “Learning sensor multiplexing design through back- propagation,” inAdv. Neural Inf. Process. Syst., vol. 29, 2016, p. 3081

    2016

  28. [36]

    Convolutional neural networks that teach microscopes how to image,

    R. Horstmeyeret al., “Convolutional neural networks that teach microscopes how to image,”arXiv:1709.07223, 2017

    2017 arXiv

  29. [37]

    Physics-based learned design: Optimized coded- illumination for quantitative phase imaging,

    M. R. Kellmanet al., “Physics-based learned design: Optimized coded- illumination for quantitative phase imaging,”IEEE Trans. Comput. Imaging, vol. 5, no. 3, pp. 344–353, 2019

    2019

  30. [38]

    End-to-end optimization of optics and image processing for achromatic extended depth of field and super-resolution imaging,

    V . Sitzmannet al., “End-to-end optimization of optics and image processing for achromatic extended depth of field and super-resolution imaging,”ACM Trans. Graph., vol. 37, no. 4, pp. 1–13, 2018

    2018

  31. [39]

    Hybrid optical-electronic convolutional neural networks with optimized diffractive optics for image classification,

    J. Changet al., “Hybrid optical-electronic convolutional neural networks with optimized diffractive optics for image classification,”Sci. Rep., vol. 8, p. 12324, 2018

    2018

  32. [40]

    Learned sensing: Jointly optimized microscope hardware for accurate image classification,

    A. Muthumbiet al., “Learned sensing: Jointly optimized microscope hardware for accurate image classification,”Biomed. Opt. Exp., vol. 10, no. 12, pp. 6351–6369, 2019

    2019

  33. [41]

    End-to-end nanophotonic inverse design for imaging and polarimetry,

    Z. Linet al., “End-to-end nanophotonic inverse design for imaging and polarimetry,”Nanophotonics, vol. 10, no. 3, pp. 1177–1187, 2021

    2021

  34. [42]

    Neural nano-optics for high-quality thin lens imaging,

    E. Tsenget al., “Neural nano-optics for high-quality thin lens imaging,” Nat. Commun., vol. 12, p. 6493, 2021

    2021

  35. [43]

    End-to-end optimization of metasurfaces for imaging with compressed sensing,

    G. Aryaet al., “End-to-end optimization of metasurfaces for imaging with compressed sensing,”ACS Photonics, vol. 11, pp. 2077–2087, 2024

  36. [44]

    Learning beamforming in ultrasound imaging,

    S. Vedulaet al., “Learning beamforming in ultrasound imaging,” inProc. Mach. Learn. Res., vol. 102, 2019, pp. 493–511

    2019

  37. [45]

    Channel estimation via tensor decomposition for dynamic metasurface antennas with known mutual coupling: Algorithms and experiments,

    J. Tapieet al., “Channel estimation via tensor decomposition for dynamic metasurface antennas with known mutual coupling: Algorithms and experiments,”arXiv:2603.19155, 2026

    2026

  38. [46]

    Efficient complementary metamaterial element for waveguide- fed metasurface antennas,

    I. Yooet al., “Efficient complementary metamaterial element for waveguide- fed metasurface antennas,”Opt. Exp., vol. 24, no. 25, p. 28 686, 2016

    2016

Pith tools

Reviewed July 2, 2026 · model on record in the stance chip above.