Pith. sign in

REVIEW 4 major objections 5 minor 4 cited by

Characterization of Indoor RIS-Assisted Channels at 304 GHz: Experimental Measurements, Challenges, and Future Directions

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper claims that a closed-form near-field correction predicts measured 304 GHz RIS path gains down to 0.25 m, after a constant 5.5 dB systematic offset is applied.

desk verdict New 304 GHz RIS measurements are valuable, but the headline near-field validation rests on a 5.5 dB offset from an unreviewed companion paper, so treat the quantitative agreement as conditional. read the letter →

arxiv 2412.07359 v1 pith:3SYE74TF submitted 2024-12-10 cs.IT cs.ETmath.IT

George C. Alexandropoulos , Bo Kum Jung , Panagiotis Gavriilidis , Sérgio Matos , Lorenz H. W. Loeser , Varvara Elesina , Antonio Clemente , Raffaele D'Errico
show 2 more authors
Luís M. Pessoa Thomas Kürner
This is my paper · ORCID
classification cs.ITcs.ETmath.IT
keywords reconfigurableintelligentsurfacesTHzcommunications304GHzchannelsoundingpathlossmodelingnear-fieldbeamformingpowerangularprofileindoorpropagation
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

This paper reports indoor channel measurements at 304 GHz in which the transmitter–receiver link is completed through a reconfigurable intelligent surface (RIS), using three static RIS prototypes with 1-, 2-, and 3-bit phase quantization. Its central claim is that the measured Tx-RIS-Rx path gain follows the near-/far-field analytical formula (2), which multiplies the standard bistatic radar equation by a beamforming-error factor $K$, with close agreement for RIS-receiver distances down to 0.25 m. The agreement is obtained after a constant 5.5 dB offset is applied to the measured data to absorb systematic errors such as sounder instability, misalignment, and fabrication imperfections. The authors argue this validates a recently proposed near-field beamforming model for sub-THz RIS links and shows that a 3-bit RIS reshapes the indoor multipath profile, reducing leakage compared with a flat reflector. The significance is that system designers could rely on a simple closed-form path-gain model instead of full-wave simulation or ray tracing when planning indoor THz coverage assisted by RISs.

What carries the argument

The central object is the near-/far-field path-gain formula (2), $P_{\mathrm{nf,ff}} = K^2(d_2,30^\circ)\,P_{\mathrm{ff}}$, which multiplies the conventional bistatic-radar path gain by the squared beamforming-error factor $K(d_2,30^\circ)$ taken from Lemma 1 of the authors' earlier work. $K$ quantifies the gain loss when the receiver at distance $d_2$ and angle $30^\circ$ sits inside the depth of focus of the RIS rather than in its far field. The argument is carried by three fabricated static RISs with $100\times100$ half-wavelength-spaced unit cells and 1-, 2-, and 3-bit phase quantization, whose measured RCS values (Table I) provide the $\sigma_{\mathrm{RIS}}$ entering the far-field formula. The $K$ correction is the load-bearing piece: it converts a far-field model that overestimates short-distance gain into one that tracks the measurements down to $0.25$ m.

What would settle it

Run the same setup with an independent absolute calibration of the 304 GHz channel sounder (for example, a direct cable-back-to-back measurement or a known free-space path between two horn antennas) and re-measure the Tx-RIS-Rx path gain without applying any offset; if the measured curve between $d_2 = 0.25$ m and $1.15$ m deviates from formula (2) by more than the reported 2.8 dB, the validation collapses.

Watch

Extended reading notes

Core claim

The paper's central claim is that the end-to-end gain of an indoor Tx-RIS-Rx link at 304 GHz, measured over RIS-receiver distances from 0.2 to 10 m, is predicted by the near-/far-field formula $P_{\mathrm{nf,ff}} = K^2(d_2,30^\circ)\,P_{\mathrm{ff}}$, where $P_{\mathrm{ff}} = \sigma_{\mathrm{RIS}}\lambda^2/((4\pi)^3 d_1^2 d_2^2)$ is the standard bistatic radar equation built from measured RCS values, and $K$ is a near-field beamforming-error factor computed from the RIS's effective aperture. The authors report that, once a constant 5.5 dB correction is applied to the measurements to absorb systematic errors, the formula matches the data for $d_2 \geq 0.25$ m, whereas the far-field-only formula overestimates the gain at short distances; at $d_2 = 0.41$ m the measured deviation from the far-field prediction is about 2.8 dB, close to the predicted 3 dB. The paper further claims that power angular profile measurements show the 3-bit RIS concentrating power in its designed $30^\circ$ beam and reducing scattered multipath relative to a flat conducting reflector, which it reads as evidence that phase-quantized RISs can shape indoor sub-THz channels.

Load-bearing premise

The load-bearing premise is that a single constant 5.5 dB correction, taken from a separate measurement study rather than an independent calibration of this setup, fully accounts for all systematic errors in the measured path gains; without that offset, the claimed agreement between the data and formula (2) does not hold.

Editorial extensions

If this is right

  • Indoor THz link budgets can use the closed-form formula (2) with effective aperture instead of full-wave simulation for RIS-receiver distances above about 0.25 m.
  • Below the depth-of-focus distance $r_{\mathrm{DF}} \approx 1.15$ m (refined to 0.41 m with effective aperture), the far-field formula (1) overestimates the path gain, so near-field-aware models are needed for short-range RIS deployments.
  • Higher phase resolution yields sharper beam shaping: the 3-bit RIS shows a 15 dB ratio between main and mirror beams and produces the strongest measured multipath component, while the 1-bit RIS splits power equally and lets a wall reflection dominate.
  • An optimized non-specular RIS reduces unwanted scattering compared with a flat conducting reflector, so the RIS phase profile can shape the indoor angular spread rather than merely adding reflections.

Reading between the lines

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

  • If the 5.5 dB offset is confirmed by an independent calibration, the same $K$-correction should scale with effective aperture to other sub-THz RIS sizes and frequencies, giving a parameter-light way to predict short-range RIS gains.
  • The angular-profile data suggest a design trade-off: a flat conducting surface increases spatial multipath diversity, while a focused RIS suppresses alternative paths; which one helps depends on whether the goal is spatial multiplexing or interference control.
  • A sharper test of the near-field model would be a full angular sweep at a fixed $d_2$, comparing measured and simulated patterns in both angle and distance, since the current agreement is demonstrated mainly along the designed $30^\circ$ cut.
  • Independent replication with a calibrated sounder and a different RIS fabrication run would settle whether the 5.5 dB correction and the near-field factor $K$ are general or specific to this prototype set.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. This paper reports an indoor measurement campaign at 304 GHz using three static RISs with 1-, 2-, and 3-bit phase quantization, each with 100x100 half-wavelength unit cells designed for a non-specular reflection at 30 degrees. In Setup 1, the measured Tx-RIS-Rx path gain is compared with the far-field radar equation (1) and with a near-/far-field corrected formula (2) that uses the RCS values of Table I and the correction factor K from reference [9]; after applying a constant 5.5 dB correction attributed to reference [12], the paper claims excellent agreement for d2 >= 0.25 m and concludes that the near-field analysis in [9] is validated. In Setup 2, PAP measurements with the 1-bit RIS, the 3-bit RIS, and a PEC backflip are compared qualitatively, and the paper closes with a list of challenges and research directions for THz RIS design and channel modeling.

Significance. If the validation claim holds, this would be one of the first experimental confirmations of a near-field beamforming-error model for RIS-assisted links at sub-THz frequencies. The paper has clear strengths: the three hardware prototypes are described in detail, the RCS values and aperture efficiencies are reported, full-wave simulations are used to support the design, and the PAP comparison against a PEC baseline is a useful experimental reference. The main limitation is that the quantitative validation of Eq. (2) depends on a 5.5 dB constant correction taken from an under-review companion paper, with no error bars or uncorrected data shown. Because the central claim rests on this calibration step, the paper is not yet a self-contained quantitative validation, although the measurement campaign and design contributions remain valuable.

major comments (4)
  1. [Section III-B, Eq. (2), Fig. 4] The central validation is conditional on the 5.5 dB correction from reference [12]. This correction is not derived from an independent calibration in the present manuscript, the uncorrected measurements are not shown, and no error bars or noise floor information are provided. Since the claimed 'excellent agreement' concerns absolute path-gain levels, the reader cannot test whether Eq. (2) is actually validated. Please show the raw (uncorrected) data, quantify measurement uncertainty, and provide a sensitivity analysis with respect to the offset value (e.g., for offsets in the range 3-8 dB).
  2. [Section III-B, Fig. 4] A constant dB shift cannot absorb distance-dependent systematic errors such as pointing misalignment, beam squint, or near-field multipath contributions. The distinguishing feature of the data is the growing gap between Eq. (1) and measurements at small d2; if any of these errors also depends on d2, the constant offset could either create or mask the near-field effect attributed to K. Please provide evidence that the 5.5 dB offset is constant over the entire d2 range and quantify the residual distance-dependent uncertainty before claiming that the near-field formula is validated.
  3. [Section III-B and Refs. [9], [12]] The near-field correction K is taken from reference [9], authored by two co-authors of this manuscript, and the 5.5 dB offset is taken from reference [12], with overlapping authorship. This does not by itself invalidate the comparison, but it makes the validation of Eq. (2) partially self-referential. A stronger test would be a comparison against an independently computed full-wave prediction of the Tx-RIS-Rx path gain, or at least a quantitative uncertainty budget for K and for the offset. Please add such a benchmark or an explicit statement of the independence of the correction.
  4. [Section III-C, Fig. 5] The PAP analysis is qualitative. In particular, the claim that the 3-bit RIS 'results in larger spatial diversity' than the 1-bit RIS is not supported by a quantitative metric. If the PAP comparison is intended as a characterization result, please report angular spread values, the number of MPCs above a defined threshold, and the associated measurement uncertainty, rather than relying only on visual inspection of Fig. 5.
minor comments (5)
  1. [Section II, Fig. 2] The text says the simulated radiation patterns are almost identical for r > 4 m, but later states they are almost identical for r > 1 m; please reconcile these statements and indicate the applicable distance on Fig. 2.
  2. [Fig. 4] The x-axis tick labels appear corrupted or duplicated (e.g., '0 0.2 0.4 1 1 1.5 2'); please correct the axis and consider marking the key distances d2 = 1.15 m and d2 = 0.25 m explicitly.
  3. [References [9], [11], [12]] Several references directly used in the quantitative analysis are under review, and no preprint is given for [12], which provides the 5.5 dB correction. Since the reader cannot verify the correction, please include a public preprint or an appendix documenting the measurement conditions and calibration methodology of [12].
  4. [Section III-C, Fig. 5] The PAP subfigures use circle sizes to indicate relative MPC strength, but the absolute normalization and the detection threshold for what counts as an MPC are not stated; please specify them so the comparison between the RIS and PEC cases is reproducible.
  5. [Abstract and Section IV-B] The abstract uses 'field-trial measurements' while the experiments are conducted in a lecture room; 'indoor measurement campaign' would be more precise. Also, in Section IV-B the statement that the PEC yields 'more channel paths' should be quantified with the MPC detection threshold just mentioned.

Circularity Check

1 steps flagged · score 4.0 of 10

The validation of the near-field formula is partly self-referential: the 5.5 dB correction comes from a companion paper under review [12] and the K factor from co-authored [9], so the absolute agreement is conditioned on self-cited inputs, though the distance-dependent shape is measurement-driven.

  1. self citation load bearing [Section III-B, Eqs. (1)-(2), Fig. 4]
    "To account for the systematic measurement errors across all considered d2 values ... a constant correction of 5.5 dB has been applied to the measured results [12]. ... the latter does not appear with the predicted values from the near-/far-field formula (2), where excellent agreement with measured data is exhibited for d2 ≥ 0.25 m, thus, validating the near-field analysis in [9]."

    The validation of the near-field model in [9] is performed using the same authors' model: K in (2) is taken from [9], and the measured data are shifted by a 5.5 dB constant taken from [12], a companion paper under review with overlapping authorship. The 'excellent agreement' is thus not an a priori prediction but a comparison made after applying a self-cited, unreviewed calibration constant. The distance-dependent gap between (1) and (2) is not created by the constant offset and reflects the measurements, so the circularity is partial.

full rationale

The paper's strongest claim is that Eq. (2) is validated by 'excellent agreement' with measured path gain for d2 ≥ 0.25 m. That validation depends on two non-independent inputs: the near-field correction K from [9] (co-authored by two of the present authors) and a 5.5 dB measurement correction from [12] (a companion paper under review with overlapping authors). The absolute-level agreement is therefore conditioned on a self-citation chain rather than on an independent calibration trace. However, the qualitative advantage of (2) over (1) is a distance-dependent shape effect—the growing gap at short d2—that a constant dB offset cannot manufacture, and this shape is present in the measured data. Thus the central qualitative conclusion has independent empirical content, and the circularity is moderate rather than total. No self-definitional or renaming circularity was found.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The paper's central validation depends on the near-field model from the authors' own prior work and on a fitted 5.5 dB correction to the measurements. These are the main debts. The RCS values come from full-wave simulations, which are independent of the measurements but not publicly available.

free parameters (1)
  • Constant path gain correction = 5.5 dB
    Applied to all measured path gains in Setup 1 (Section III-B) to compensate for systematic errors; referenced to the companion paper [12], but not derived or calibrated in this paper. Without it, the agreement with the models would be much worse.
assumptions (5)
  • domain assumption The free-space radar equation (1) with the RCS values from Table I correctly models the RIS far-field path gain.
    The paper uses (1) as the baseline far-field model; it assumes the RIS is a point scatterer with a single RCS value.
  • domain assumption The near-field beamforming error model K(d2, 30°) from [9, Lemma 1] is valid for the designed static RISs.
    The paper adopts this model (written by two co-authors) to correct the far-field formula; its validity is what the paper claims to validate, making this assumption load-bearing.
  • domain assumption A constant 5.5 dB correction accounts for all systematic measurement errors and is independent of d2.
    Applied to all measured values in Setup 1; no calibration details are given in this paper, and the offset is a free parameter.
  • domain assumption The horn antennas have the stated gain of 26.4 dBi and HPBW of 8.5° throughout the band.
    Used in the path gain calculations; no antenna pattern corrections are applied.
  • domain assumption The dominant multipath components identified in Setup 2 capture the environment's essential propagation behavior.
    The PAP analysis selects four dominant MPCs; other paths are neglected or attributed to noise.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Characterization of Indoor RIS-Assisted Channels at 304 GHz: Experimental Measurements, Challenges, and Future Directions." pith.science (2026). https://pith.science/paper/3SYE74TF

@misc{pith2026241207359,
  author       = {Pith},
  title        = {Pith review of: Characterization of Indoor RIS-Assisted Channels at 304 GHz: Experimental Measurements, Challenges, and Future Directions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3SYE74TF}},
  note         = {Machine review of arXiv:2412.07359}
}
abstract

Reconfigurable Intelligent Surfaces (RISs) are expected to play a pivotal role in future indoor ultra high data rate wireless communications as well as highly accurate three-dimensional localization and sensing, mainly due to their capability to provide flexible, cost- and power-efficient coverage extension, even under blockage conditions. However, when considering beyond millimeter wave frequencies where there exists GHz-level available bandwidth, realistic models of indoor RIS-parameterized channels verified by field-trial measurements are unavailable. In this article, we first present and characterize three RIS prototypes with $100\times100$ unit cells of half-wavelength inter-cell spacing, which were optimized to offer a specific non-specular reflection with $1$-, $2$-, and $3$-bit phase quantization at $304$ GHz. The designed static RISs were considered in an indoor channel measurement campaign carried out with a $304$ GHz channel sounder. Channel measurements for two setups, one focusing on the transmitter-RIS-receiver path gain and the other on the angular spread of multipath components, are presented and compared with both state-of-the-art theoretical models as well as full-wave simulation results. The article is concluded with a list of challenges and research directions for RIS design and modeling of RIS-parameterized channels at THz frequencies.

Figures

Figures reproduced from arXiv: 2412.07359 by the authors.

Figure 1
Figure 1. The fabricated static RISs at 304 GHz with 100 × 100 unit cells of half-wavelength inter-cell spacing and 1-, 2-, and 3-bit phase quantization designed to offer a non-specular reflection from a normal incidence to a 30◦ outgoing plane. dense indoor environments, channel models that accurately predict the behavior resulting from all possible RIS phase configurations, are needed. Current literature though often relies… view at source ↗
Figure 2
Figure 2. Normalized radiation pattern of the designed [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Schematic view of the indoor channel measurements setups at [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Comparison of the measured and theoretical (both [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Power angular profile measurements for the Setup 2 illustrated in Fig. 3b considering three different cases for the RIS. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Evaluation of Switching Technologies for Reflective and Transmissive RISs at Sub-THz Frequencies

    cs.ET 2025-04 conditional novelty 5.0 of 10

    The authors present simulated and one measured D-band RIS unit-cell designs using five switching technologies, reporting insertion losses and beam steering patterns.

  2. Reconfigurable Intelligent Surfaces for 6G and Beyond: A Comprehensive Survey from Theory to Deployment

    eess.SP 2025-06 accept novelty 4.0 of 10

    A comprehensive survey of RIS for 6G that integrates use cases, control mechanisms, channel sounding, channel estimation, and standardization and industry perspectives.

  3. Indoor Channel Characterization with Extremely Large Reconfigurable Intelligent Surfaces at $300$ GHz

    cs.IT 2025-01 conditional novelty 4.0 of 10

    A 100x100 two-bit RIS at 304 GHz can be modeled as three discrete rays, and its far-field approximation holds at roughly one fifth of the textbook far-field distance.

  4. RIS-Assisted MIMO CV-QKD at THz Frequencies: Channel Estimation and SKR Analysis

    cs.IT 2024-12 conditional novelty 4.0 of 10

    A closed-form secret key rate expression is derived for RIS-assisted MIMO continuous-variable QKD at THz with least-squares channel estimation and collective Gaussian eavesdropping.

Reference graph

Works this paper leans on

15 extracted references · 12 canonical work pages · cited by 4 Pith papers

  1. [12]

    Channel measurements involving passive RIS at 300 GHz,

    L. H. W. Loeser, S. Matos, and T. Kürner, “Channel measurements involving passive RIS at 300 GHz,” in IEEE Eur. Conf. Antennas Propag., Stockholm, Sweden, 2025, under review

    work page 2025

  2. [9]

    Near-field beam tracking with extremely large dynamic metasurface antennas,

    P. Gavriilidis and G. C. Alexandropoulos, “Near-field beam tracking with extremely large dynamic metasurface antennas,” IEEE Trans. Wireless Commun. , 2024, revision under review. [Online]. Available: https://arxiv.org/abs/2406.01488

    arXiv 2024

  3. [1]

    Terahertz modeling (THz); identification of use cases for THz communication systems,

    “Terahertz modeling (THz); identification of use cases for THz communication systems,” ETSI GR THz 001, ETSI, Jan. 2024

    work page 2024

  4. [2]

    Mobile terahertz communication and sensing systems: A future look,

    J. M. Jornet, H. Elayan, T. Nagatsuma, M. Juntti, E. T. R. Pinto, T. Kürner, H. Guerboukha, D. M. Mittleman, and E. Knightly, “Mobile terahertz communication and sensing systems: A future look,” IEEE Veh. Technol. Mag., early access, 2024

    work page 2024

  5. [3]

    Reconfigurable intelligent surfaces for THz: Hardware impairments and switching technologies,

    S. Matos, Y . Ma, Q. Luo, J. Deuermeier, L. Lucci, P. Gavriilidis, A. Kiazadeh, V . Lain-Rubio, T. D. Phan, P. J. Soh, A. Clemente, L. M. Pessoa, and G. C. Alexandropoulos, “Reconfigurable intelligent surfaces for THz: Hardware impairments and switching technologies,” in Proc. IEEE Int. Conf. Electromagn. Adv. Appl. , Lisbon, Portugal, 2024

    work page 2024

  6. [4]

    Reconfigurable intelligent surfaces for 6G: Emerging hardware architectures, applications, and open challenges,

    E. Basar, G. C. Alexandropoulos, Y . Liu, Q. Wu, S. Jin, C. Yuen, O. A. Dobre, and R. Schober, “Reconfigurable intelligent surfaces for 6G: Emerging hardware architectures, applications, and open challenges,” IEEE Veh. Technol. Mag., vol. 19, no. 3, pp. 27–47, 2024

    2024

  7. [5]

    Reconfigurable intelligent surfaces for THz: Hardware design and signal processing challenges,

    G. C. Alexandropoulos, A. Clemente, S. Matos, R. Husbands, S. Ahearne, Q. Luo, V . Lain-Rubio, T. Kürner, and L. M. Pessoa, “Reconfigurable intelligent surfaces for THz: Hardware design and signal processing challenges,” in Proc. IEEE Eur. Conf. Antennas Propag., Glasgow, Scotland, 2024

    work page 2024

  8. [6]

    Performance analysis of THz backhaul links assisted by reconfigurable intelligent surfaces,

    B. K. Jung and T. Kürner, “Performance analysis of THz backhaul links assisted by reconfigurable intelligent surfaces,” in Proc. Eur. Conf. Antennas Prop., Glasgow, Scotland, 2024

    work page 2024

Show all 15 references
  1. [7]

    Near- field communications: A tutorial review,

    Y . Liu, Z. Wang, J. Xu, C. Ouyang, X. Mu, and R. Schober, “Near- field communications: A tutorial review,” IEEE Open J. Commun. Soc. , vol. 4, pp. 1999–2049, 2023

    1999

  2. [8]

    A primer on near-field beamforming for arrays and reconfigurable intelligent surfaces,

    E. Björnson, O. T. Demir, and L. Sanguinetti, “A primer on near-field beamforming for arrays and reconfigurable intelligent surfaces,” in Proc. Asilomar Conf. Signals Sys. Comp. , Pacific Grove, USA, 2021

    2021

  3. [10]

    Hybrid channel model for low terahertz links in a data center,

    J. M. Eckhardt, T. Doeker, and T. Kürner, “Hybrid channel model for low terahertz links in a data center,” IEEE Open J. Commun. Society , vol. 5, pp. 4731–4745, 2024

    2024

  4. [11]

    Double directional channel measurements at 300 GHz in indoor environments enhanced by reconfigurable intelligent surfaces,

    B. K. Jung, J. M. Eckhardt, V . V . Elesina, and T. Kürner, “Double directional channel measurements at 300 GHz in indoor environments enhanced by reconfigurable intelligent surfaces,” in Eur. Conf. Antennas Propag., Stockholm, Sweden, 2025, under review

    2025

  5. [13]

    Wideband high-gain transmitarray antenna for point-to-point communications at 300 GHz,

    O. Koutsos, F. F. Manzillo, A. Clemente, and R. Sauleau, “Wideband high-gain transmitarray antenna for point-to-point communications at 300 GHz,” in Proc. Eur. Microw. Conf., London, UK, 2022

    2022

  6. [14]

    The far-/near- field beam squint and solutions for THz intelligent reflecting surface communications,

    W. Hao, X. You, F. Zhou, Z. Chu, G. Sun, and P. Xiao, “The far-/near- field beam squint and solutions for THz intelligent reflecting surface communications,” IEEE Trans. Veh. Technol., vol. 72, no. 8, pp. 10 107– 10 118, 2023

    2023

  7. [15]

    On the detection of non-cooperative RISs: Scan B-testing via deep support vector data description,

    G. Stamatelis, P. Gavriilidis, A. Fakhreddine, and G. Alexandropoulos, “On the detection of non-cooperative RISs: Scan B-testing via deep support vector data description,” in IEEE Int. Conf. Commun. , Montreal, Canada, 2025, under review. [Online]. Available: https: //arxiv.or...

    2025 arXiv

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.