RIS Size Determination Across Frequencies and Deployment Scenarios: A Simulation-Based Study

Ahmet Faruk Coskun; Emre Arslan

REVIEW 3 major objections 5 minor 14 references

RIS Size Determination Across Frequencies and Deployment Scenarios: A Simulation-Based Study

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read RIS panel size must be tuned per deployment, not chosen once for all networks.

desk verdict A useful engineering study of RIS sizing across 16 use cases that is honest about its limitation, but the fixed free-space path-loss exponent makes the quantitative sizing tables unreliable for NLoS and long-range deployments. read the letter →

arxiv 2506.06528 v1 pith:QUQBLNDO submitted 2025-06-06 eess.SP cs.SYeess.SY

Emre Arslan , Ahmet Faruk Coskun This is my paper

classification eess.SPcs.SYeess.SY

keywords reconfigurableintelligentsurfacesRISsizingpath-lossmodellingsignal-to-noiseratiooutageprobabilitysub-6GHzFR3mmWave

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 tries to establish that there is no universal RIS panel size: the number of reflecting elements needed to meet a signal-to-noise target depends on carrier frequency, BS-RIS-UE geometry, and environment. The authors build a scattering-aware simulation of RIS-assisted links, define sixteen realistic use cases spanning home WiFi, sub-6 GHz urban microcells, FR3 macrocells, and mmWave indoor/outdoor settings, and compute for each the minimum RIS dimensions that reach 5, 10, 20, and 30 dB average SNR. The practical payoff is a sizing table a network planner could use as a first-order purchasing guide. If the paper is right, operators should stop asking for a single standard RIS and instead select panel dimensions per deployment class.

What carries the argument

The load-bearing machinery is the RIS-assisted signal model introduced in [6]: an instant-by-instant bi-static scattered field in which each of the $N_h \times N_v$ unit cells contributes with a complex weight $\rho_{kl} e^{-j\Psi_{kl}}$ and a delay, so the surface's angular scattering behavior enters the received power directly. The authors combine this model with a flexible simulation template (BS, RIS, and UE coordinates, heights, and distances), a large pool of random realizations per use case, and the resulting probability densities of scattered power and SNR. That machinery is what converts abstract RIS gain into concrete statements about how many elements are needed to keep outage probability below a boundary or SNR above a threshold. It also lets the RIS altitude be varied, showing that lower mounting makes performance more sensitive to where the UE is located.

What would settle it

Measure the SNR CDF at a dense urban sub-6 GHz microcell site using the Table II parameter distribution and a panel sized by Fig. 5 for the 10 dB target; if the empirical median SNR falls below 10 dB, the fixed path-loss exponent of 2 is too optimistic and the table needs environment-specific recalibration.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that RIS-aided performance scales with surface aperture but the marginal gain saturates at scenario-dependent sizes, so the minimum panel that meets a target SNR varies widely across use cases. The simulation uses a bi-static scattering signal model that accounts for the angular response of the surface; it is validated against 26 GHz outdoor trial measurements, where the simulated SNR trend follows the measured curve when an empirical path-loss exponent of 1.785 is used. Applying the model to the sixteen use cases shows, for example, that indoor scenarios need only modest panels while rural FR3 and mmWave macro scenarios require very large surfaces (beyond $40\times40$ elements) to reach the same targets. The paper presents these as minimum RIS sizes for 5, 10, 20, and 30 dB SNR thresholds in a single summary figure.

Load-bearing premise

The sizing numbers all rely on one generic path-loss formula with a fixed exponent of 2; real indoor, urban, and rural channels behave differently, so the minimum sizes in the summary figure could shift substantially.

Editorial extensions

If this is right

For a given SNR target, the summary figure in the paper gives the minimum RIS dimensions for each use case, so an operator can translate a coverage requirement into a panel purchase specification.
Outage probability drops sharply once the RIS exceeds roughly $20\times20$ elements, so for indoor and short-range outdoor cases extra elements buy little reliability; larger apertures pay off mainly in high-frequency, long-distance, sparse-scattering deployments.
RIS mounting height matters as much as panel size: at lower altitudes the SNR at the UE varies strongly with bearing, while higher mounting yields more uniform coverage and less dependence on exact UE placement.
The validation at 26 GHz behaves more closely with an empirical path-loss exponent of 1.785 than with a free-space exponent of 2, so the simulator's default 2 may under- or over-estimate required sizes depending on the environment.
There is no universal panel; an RIS sized for a home WiFi hotspot is far too small for a rural FR3 macrocell, and a rural panel would be wasteful indoors.

Reading between the lines

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

If the $20\times20$ saturation trend holds, a sensible product strategy would be to standardize on one moderate panel for indoor and short-range links and a separate large-aperture product for rural macro, rather than a single SKU.
The fixed path-loss exponent of 2 is the main lever: replacing it with environment-specific exponents (as the paper's conclusions flag) could change the Fig. 5 sizes by several dB per decade of distance, so the table should be read as a starting point rather than a final specification.
The same simulation machinery could be inverted into a sizing calculator that takes frequency, deployment class, target SNR, and outage bound as inputs and returns a recommended panel area, making the methodology directly usable in network-planning tools.
A natural next experiment is to repeat the validation at 3.5 GHz or in an indoor factory; if the SNR trend there deviates from the model, the angular scattering assumptions would need refinement before the sizing table is used for those bands.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Read it. The paper does something the field still needs: it converts the RIS sizing question from a theoretical exercise into a set of concrete numbers across frequencies and deployment scenarios. The 16 use case tables and the minimum-size summary in Fig. 5 are useful as a first-order planning aid. The authors also deserve credit for testing the simulator against a real ZTE trial and for using an externally measured PLE (1.785) in the validation, even though that value comes from an indoor office channel.

The soft spot is the fixed PLE=2. The stress-test note is right: the sizing results use free-space path loss for every use case, including rural macro NLoS at 2 km. The paper's own validation shows PLE=1.785 fits the mmWave trial better than PLE=2, yet the use cases all default to PLE=2. For sub-6 and FR3 NLoS at long range, effective PLEs are usually well above 2 (3GPP TR 38.901). At 2 km, the difference between PLE=2 and PLE=3.5 is tens of dB, and the required RIS aperture grows roughly with that deficit, so the element counts in Fig. 5 can be too small by an order of magnitude for those long-range cases. The authors admit the fixed PLE in the conclusion, but they don't caveat Fig. 5 itself, which is the operational takeaway.

A separate medium issue: the 'minimum size' selection procedure is under-specified. It isn't clear whether the SNR thresholds apply to the mean or to a percentile of the distribution, and the outage threshold is never defined. The use-case parameter tables give discrete sets, but the sampling distribution isn't stated. These gaps make the numbers hard to reproduce.

The validation is only qualitative: visual trend agreement, no error metric, and one scenario at 26 GHz over distances up to 205 m. That supports the model's qualitative behavior but does not validate transferability to sub-6 GHz or FR3 or to 2 km links.

The central qualitative claim—no one-size-fits-all RIS—is robust and won't change with better path loss. But the quantitative minimum sizes in Fig. 5 should not be used for procurement decisions until the authors rerun with environment-specific PLEs and publish code/data.

I'd send this to review rather than desk reject. It's a competent engineering study with practical value; the PLE issue is fixable and already acknowledged. A referee can push for the rerun and for a better specification of the method. If that happens, it becomes a genuinely useful design guideline. I'd bring it to a reading group only if people want to discuss what 'validation' means here.

Referee Report

3 major / 5 minor

Summary. The paper proposes a simulation-based methodology for determining the minimum RIS size (number of unit cells and physical aperture) required to meet target average SNR thresholds across 16 deployment use cases spanning sub-6 GHz, FR3, mmWave, indoor, urban, and rural scenarios. The signal model is taken from the authors' earlier work [6] and includes angular scattering behavior and finite phase states. The simulator is checked against a single ZTE/Turkcell 26 GHz outdoor trial, and then applied to a parameterized set of use cases whose geometries and link budgets are informed by 3GPP TR 38.901 and related references. The results are reported as PDFs of received power and SNR for three representative use cases (UC-5, UC-7, UC-16), outage-probability curves versus RIS size, and a summary figure (Fig. 5) giving minimum RIS sizes for SNR thresholds of 5, 10, 20, and 30 dB. The central qualitative conclusion is that RIS dimensions must be tailored to frequency, geometry, and environment, and that no one-size-fits-all RIS size exists.

Significance. If the quantitative sizing results were reliable, the paper would provide a genuinely useful first-order design guideline for operators and vendors, and the use-case tables would be a convenient starting point for standardization discussions. The qualitative conclusion that no single RIS size works across all deployments is very likely correct and is supported by the simulations. The paper has tangible strengths: it uses a forward simulation model rather than fitting the conclusion to the data, it anchors deployment parameters in recognized 3GPP technical reports, and it includes an external field-trial data set for validation. However, the operational value of the paper rests on the quantitative minimum sizes in Fig. 5, and those numbers depend on a fixed free-space path-loss exponent and on an averaging procedure over parameter sets whose probability weights are not specified. The single 26 GHz validation is best interpreted as a sanity check rather than as proof of accuracy across all 16 use cases. For these reasons, the contribution is useful but the quantitative claims are not yet established at the level required for an archival journal recommendation.

major comments (3)

[Section IV and Section II-C] The quantitative sizing results in Fig. 5 are computed with a generic Friis model using a fixed path-loss exponent of 2 in every use case, as explicitly acknowledged in Section IV. This is not a conservative assumption in general: for the long rural and macro links in Tables II and III (e.g., UC-3, UC-6, UC-9, UC-16, with BS-RIS and RIS-UE distances up to 2000 m), the 3GPP TR 38.901 models cited in Section II-E imply effective path-loss exponents well above 2, particularly for NLoS conditions. Since the required RIS aperture increases rapidly with path-loss deficit, the minimum element counts and even the ranking of use cases in Fig. 5 could change substantially under scenario-specific PLEs. The validation in Section II-C, which at most supports a PLE of about 1.785 at 26 GHz over distances up to 205 m with LoS between RIS and UE, does not calibrate the sub-6 GHz, FR3, indoor, or long-range outdoor cases. BecauseFig. 5 is the central operational deliverable, this fixed-PLE assumption is load-bearing and must be addressed, for example by a sensitivity analysis over PLE values or by using the TR 38.901 scenario-specific path-loss models directly.
[Section II-C] The external validation is based on a single scenario: a 26 GHz outdoor deployment with LoS between the RIS and the UE, distances up to 205 m, and only qualitative comparison of trends. The paper reports no quantitative error metric (e.g., RMSE, bias, or scatter plot of predicted versus measured SNR), and the agreement is described only as 'consistency in the main trend.' Moreover, the empirical PLE of 1.785 is taken from [8], which is an indoor-office channel model, and is not measured in the outdoor trial itself. This validation is therefore too narrow to support the statement that the simulator 'accurately models RIS-assisted propagation scenarios' across sub-6 GHz, FR3, indoor, NLoS, and long-range rural use cases. I recommend either restricting the accuracy claim to the validated geometry or providing additional validation data for at least one non-mmWave and one indoor indoor scenario.
[Section II-E and Tables II-IV] The use-case parameters are presented only as discrete lists (e.g., hBS in {10,15,20,30}, dBS-RIS in {500,1000,1500,2000}), with no joint distribution, weights, or sampling rule stated. The methodology states that a 'large pool' of data is generated and that PDFs of power and SNR are extracted, but it does not specify how many realizations are drawn from each parameter combination, whether the parameters are uniform or weighted, or how the averaging over the use case is performed. Consequently, the PDFs in Fig. 3, the outage-probability curves in Fig. 4, and the minimum sizes in Fig. 5 are not uniquely defined by the manuscript. The averaging rule matters directly for the quantitative conclusions, since Fig. 5 aggregates across altitudes, distances, and UE heights. Please specify the exact probability model over Tables II-IV and the algorithm used to extract 'minimum RIS size' for a given SNR threshold.

minor comments (5)

[Section II-E] The text refers to 'preliminary results given in Fig. 2-(c)', but Fig. 2 contains only subpanels (a) and (b); the reference should be corrected or the missing subpanel should be added.
[Section II-B] The notation is inconsistent: y[n] is introduced as 'the signal part of r[n]' only after µ[n] is defined in terms of y[n], and the relation between the double-sum in Eq. (1) and the instantaneous signal power µ[n] is not explicit. Define y[n] first and rewrite the SNR expression accordingly.
[Section III and Fig. 4] The outage probability plotted in Fig. 4 is never defined. Please state the outage threshold (presumably an SNR or SINR value) used to compute the OP curves.
[Throughout] Physical quantities are frequently written without a space before the unit (e.g., '40m', '160m', '205meters', '15kHz'). Please use consistent SI formatting throughout the manuscript.
[Fig. 5] The figure caption and the text should clarify whether 'RIS size' on the axis denotes the number of unit cells along each dimension, the total element count, or the physical aperture in square meters; the current description is ambiguous for an operator audience.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the sizing results are forward simulation outputs, not fitted quantities; the self-cited scattering model is grounded by an external ZTE trial and external PLE values.

full rationale

The paper's central outputs (Fig. 5 minimum RIS sizes, SNR/OP curves) are produced by forward simulation from a stated signal model and use-case parameter tables; no target quantity is used as an input to define the model. The path-loss exponents used (2 and 1.785) come from the standard free-space model and from external measurement study [8], not from fitting to the paper's own sizing results. The only author-overlapping input is the scattering model of [6], but it is not an unverified black box: Section II-C validates the simulator against an external ZTE mmWave RIS trial [7], and the same model's predictions are compared with measured SNR trends. The acknowledged fixed-PLE limitation in Section IV is a modeling restriction that affects accuracy, not a circular reduction. There is no equation in which a claimed prediction equals a fitted parameter by construction, and no uniqueness or ansatz is imported from the authors' prior work. Therefore no circular step is present.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The central claims rest on the authors' prior scattering model [6] plus a set of hand-chosen propagation and deployment parameters. The only external anchor is a single 26 GHz field trial; all other quantitative outputs follow from the assumed model and are not independently verified.

free parameters (4)

Path-loss exponent (PLE) = 2 for all use cases; 1.785 in validation from [8]
The central sizing numbers in Figs. 3-5 use PLE=2. Real channels have environment-specific exponents, so this hand-picked constant directly controls the required RIS size.
Use-case parameter sets (heights, distances, powers) = 16 configurations in Tables II-IV
These discrete sets are selected from standards and operator experience, but they are not measured for the modeled deployments. The KPI distributions and minimum-size results depend on these choices.
SNR thresholds for sizing = 5, 10, 20, 30 dB in Fig. 5
These are chosen targets, not derived from service requirements. The reported minimum sizes are only meaningful relative to these thresholds.
Number of RIS phase states = KPhi=2 (1-bit)
Table I sets only two phase states. This constrains the achievable beamforming gain and affects the size estimates.

assumptions (6)

domain assumption The RIS scattering signal model of [6] is valid for all simulated frequencies (3.5, 6, 8, 27 GHz) and environments.
The paper reuses [6] without re-validating outside the 26 GHz outdoor geometry; angular scattering behavior may differ across bands and indoor/outdoor settings.
domain assumption Free-space propagation with PLE=2 holds for indoor, urban, and rural use cases.
Section IV admits the fixed-PLE limitation; this is load-bearing because Fig. 5 minimum sizes scale with path loss.
domain assumption The direct BS-UE path is perfectly obstructed in all considered use cases.
Section II-B assumes NLoS between BS and UE; if a direct path exists, RIS gains and required sizes would differ.
domain assumption The empirical PLE=1.785 from the indoor office study [8] is transferable to the outdoor 26 GHz validation scenario.
Section II-C uses this PLE to match the ZTE trial; no outdoor-specific measurement justifies it.
ad hoc to paper Discrete RIS unit cells scatter independently with no mutual coupling and with only two phase states.
Table I sets KPhi=2 and lambda/2 spacing; this simplified model underlies all KPI computations.
ad hoc to paper The single 26 GHz field trial with visual trend agreement is sufficient to assert simulator accuracy across all 16 use cases.
Section II-C compares one geometry with a qualitative fit; there is no quantitative error metric and no check for other bands.

how reviews work

0 comments

Cite this review

Pith. "Pith review of RIS Size Determination Across Frequencies and Deployment Scenarios: A Simulation-Based Study." pith.science (2026). https://pith.science/paper/QUQBLNDO

@misc{pith2026250606528,
  author       = {Pith},
  title        = {Pith review of: RIS Size Determination Across Frequencies and Deployment Scenarios: A Simulation-Based Study},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QUQBLNDO}},
  note         = {Machine review of arXiv:2506.06528}
}

read the original abstract

Despite the growing interest in the integration of reconfigurable intelligent surfaces (RIS) into next-generation wireless communications systems, a critical gap remains in understanding what the dimensions of an RIS must be to provide meaningful performance gains across realistic deployment scenarios. This paper addresses this challenge by presenting a practical and scenario-aware methodology for determining optimal RIS dimensions, tailored to specific frequency bands, environments, and use cases. Leveraging a realistic simulation model that incorporates angular scattering characteristics, practical network node locations, and propagation constraints, we evaluate the RIS-assisted performance in a diverse set of configurations. For selected use-cases, we quantify key performance indicators such as average signal-to-noise ratio and outage probability, and we demonstrate how RIS size impacts system reliability. Our findings show that RIS deployment effectiveness is highly sensitive to both physical size and geometric placement, and that there is no one-size-fits-all solution. The proposed framework, supported by detailed use case tables and validated through comprehensive simulations, offers design guidelines for operators and vendors seeking to deploy RIS in practical wireless network settings.

Figures

Figures reproduced from arXiv: 2506.06528 by the authors.

**Figure 2.** Validation scenario and results In the measurement study, the BS was mounted on a tower at an altitude of 88.7 meters and operated at a center frequency of 26 GHz. The RIS product of ZTE (i.e., Dynamic RIS 2.0) was installed 80 meters distant from the BS on a second-floor balcony at an altitude of 64 meters. The UE test locations sampled with totally 141 coordinate points (corresponding to distances up to 205 meters… view at source ↗

**Figure 4.** Average RIS-assisted signal power (left), SNR (middle) [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗

Figures from the paper (1 more)

**Figure 5.** Figure 5: Minimum RIS sizes required to achieve different SNR [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

14 extracted references · 14 canonical work pages

[6]

R. Liu, J. Dou, P. Li, J. Wu and Y. Cui, ``Simulation and Field Trial Results of Reconfigurable Intelligent Surfaces in 5G Networks," in IEEE Access, vol. 10, pp. 122786-122795, 2022

work page 2022
[8]

M. K. Simon and M.-S. Alouini, Digital communications over generalized fading channels: A unified approach to performance analysis, New York: Wiley, 2005

work page 2005
[1]

W. Saad, M. Bennis and M. Chen, ``A Vision of 6G Wireless Systems: Applications, Trends, Technologies, and Open Research Problems," IEEE Network, vol. 34, no. 3, pp. 134-142, May/June 2020

work page 2020
[2]

Basar, M

E. Basar, M. Di Renzo, J. De Rosny, M. Debbah, M. -S. Alouini and R. Zhang, ``Wireless Communications Through Reconfigurable Intelligent Surfaces," IEEE Access, vol. 7, pp. 116753-116773, 2019

work page 2019
[3]

S. W. Ellingson, ``Path Loss in Reconfigurable Intelligent Surface-Enabled Channels," 2021 IEEE 32nd Annual International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), Helsinki, Finland, 2021, pp. 829-835

work page 2021
[4]

Başar, İ

E. Başar, İ. Yıldırım and F. Kılınç, ``Indoor and Outdoor Physical Channel Modeling and Efficient Positioning for Reconfigurable Intelligent Surfaces in mmWave Bands," in IEEE Trans. on Commun., vol. 69, no. 12, pp. 8600-8611, Dec. 2021

work page 2021
[5]

Tang et al., ``Wireless communications with reconfigurable intelligent surface: Path loss modeling and experimental measurement," IEEE Trans

W. Tang et al., ``Wireless communications with reconfigurable intelligent surface: Path loss modeling and experimental measurement," IEEE Trans. Wireless Commun., vol. 20, no. 1, pp. 421–439, Jan. 2021

work page 2021
[7]

Rains, J

J. Rains, J. U. R. Kazim, A. Tukmanov, L. Zhang, Q. Abbasi and M. Imran, ``Fully-Addressable Varactor-Based Reflecting Metasurface with Dual-Linear Polarisation for Low Power Reconfigurable Intelligent Surfaces," EuCAP 2023, Florence, Italy, 2023, pp. 1-5

work page 2023

Show all 14 references

[9]

A. F. Co s kun, E. Arslan, ``A Practical Modeling Approach on the Reflectivity Characteristics of Reconfigurable Intelligent Surfaces," in IEEE, 2024 6th International Conference on Communications, Signal Processing, and their Applications (ICCSPA), 1--6, Jul. 2024

2024
[10]

[Online]

ZTE Corporation, ``Turkcell and ZTE conduct Europe’s first dynamic RIS trial in a 5G-Advanced mmWave network,” 2024. [Online]. Available: https://www.zte.com.cn/global/about/news/turkcell-and-zte-conduct-europes-first-dynamic-ris-trial-in-a-5g-advanced-mmwave-network.html

2024
[11]

Rubio, V

L. Rubio, V. M. R. Peñarrocha, M. Cabedo-Fabres, B. Bernardo-Clemente, J. Reig, H. Fernández, J. R. Pérez, R. P. Torres, L. Valle, Ó. Fernández, ``Millimeter-Wave Channel Measurements and Path Loss Characterization in a Typical Indoor Office Environment. Electronics 2023, vol....

2023 doi
[12]

Rubio, et al., ``Millimeter-Wave Channel Measurements and Path Loss Characterization in a Typical Indoor Office Environment," Electronics, vol

L. Rubio, et al., ``Millimeter-Wave Channel Measurements and Path Loss Characterization in a Typical Indoor Office Environment," Electronics, vol. 12, no. 844, 2023

2023
[13]

3rd Generation Partnership Project (3GPP), TR 38.901 v16.1.0 Study on channel model for frequencies from 0.5 to 100 GHz, Technical Report, 2020

2020
[14]

3rd Generation Partnership Project (3GPP), TR 38.873 v16.0.0, Study on Integrated Access and Backhaul (IAB), Technical Report, 2019

2019

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [6]

R. Liu, J. Dou, P. Li, J. Wu and Y. Cui, ``Simulation and Field Trial Results of Reconfigurable Intelligent Surfaces in 5G Networks," in IEEE Access, vol. 10, pp. 122786-122795, 2022

work page 2022

[2] [8]

M. K. Simon and M.-S. Alouini, Digital communications over generalized fading channels: A unified approach to performance analysis, New York: Wiley, 2005

work page 2005

[3] [1]

W. Saad, M. Bennis and M. Chen, ``A Vision of 6G Wireless Systems: Applications, Trends, Technologies, and Open Research Problems," IEEE Network, vol. 34, no. 3, pp. 134-142, May/June 2020

work page 2020

[4] [2]

Basar, M

E. Basar, M. Di Renzo, J. De Rosny, M. Debbah, M. -S. Alouini and R. Zhang, ``Wireless Communications Through Reconfigurable Intelligent Surfaces," IEEE Access, vol. 7, pp. 116753-116773, 2019

work page 2019

[5] [3]

S. W. Ellingson, ``Path Loss in Reconfigurable Intelligent Surface-Enabled Channels," 2021 IEEE 32nd Annual International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), Helsinki, Finland, 2021, pp. 829-835

work page 2021

[6] [4]

Başar, İ

E. Başar, İ. Yıldırım and F. Kılınç, ``Indoor and Outdoor Physical Channel Modeling and Efficient Positioning for Reconfigurable Intelligent Surfaces in mmWave Bands," in IEEE Trans. on Commun., vol. 69, no. 12, pp. 8600-8611, Dec. 2021

work page 2021

[7] [5]

Tang et al., ``Wireless communications with reconfigurable intelligent surface: Path loss modeling and experimental measurement," IEEE Trans

W. Tang et al., ``Wireless communications with reconfigurable intelligent surface: Path loss modeling and experimental measurement," IEEE Trans. Wireless Commun., vol. 20, no. 1, pp. 421–439, Jan. 2021

work page 2021

[8] [7]

Rains, J

J. Rains, J. U. R. Kazim, A. Tukmanov, L. Zhang, Q. Abbasi and M. Imran, ``Fully-Addressable Varactor-Based Reflecting Metasurface with Dual-Linear Polarisation for Low Power Reconfigurable Intelligent Surfaces," EuCAP 2023, Florence, Italy, 2023, pp. 1-5

work page 2023