{"id":"daa20251-003a-4a26-96c1-60cc71c325d8","arxiv_id":"2506.06528","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Minimum RIS sizes vary strongly across use cases, and no single panel size works for all 5G/6G deployments.","lead":"This paper simulates reconfigurable intelligent surface (RIS) performance across frequencies and environments to estimate the panel sizes needed for indoor, urban, and rural coverage. Its message for telecom planners is that RIS size must be chosen case by case, because required sizes vary strongly with frequency, distance, and geometry.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Fixed PLE=2 across all use cases, despite 3GPP-informed deployment parameters, likely underestimates required RIS sizes for long-range and NLoS scenarios, undermining the quantitative sizing guide in Fig. 5.","rationale":"The reader's weakest_assumption identified the fixed PLE=2 as the key vulnerability. My analysis agrees and sharpens it: the authors derive use-case parameters from 3GPP TR 38.901, yet they do not adopt 3GPP's environment-specific path-loss models. Since the paper's headline contribution is a quantitative sizing guide (Fig. 5), a propagation model that is acknowledged to be a simplification, and one that is inconsistent with the cited standards, is the single most load-bearing assumption. The qualitative conclusion (no one-size-fits-all RIS) is robust, but the operational tables are not. The concrete test of rerunning with TR 38.901 path loss would settle whether the effect is large enough to change the guidance. The reader already issued CONDITIONAL; my analysis does not shift that verdict, so UNCHANGED is appropriate. Agreement is full because the reader's weakest assumption is exactly the fixed PLE issue, though I add the 3GPP inconsistency as a compounding factor.","tokens_in":7830,"tokens_out":5929,"duration_ms":63173,"concrete_test":"Recompute the minimum RIS sizes in Fig. 5 using the 3GPP TR 38.901 scenario-specific path-loss models (UMi, UMa, RMa) for both the BS-RIS and RIS-UE links, keeping all other simulation parameters identical. If the minimum element counts for UC-3, UC-6, UC-9, and UC-16 change by more than a factor of two for any SNR threshold, the quantitative sizing guide is not robust to realistic propagation conditions and the paper's operational claims need to be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central operational claim is that Fig. 5 provides minimum RIS sizes for operators across sub-6, FR3, and mmWave deployments. However, the SNR calculations use the generic Friis path-loss model with a fixed path-loss exponent (PLE) of 2 for every use case, as explicitly acknowledged in Section IV. This is inconsistent with the paper's own use-case parameterization, which is explicitly based on 3GPP TR 38.901 (Section II-E). TR 38.901 defines scenario-specific path-loss models with effective PLEs that are generally higher than 2, especially for NLoS conditions and for RMa (e.g., RMa NLoS PLE often exceeds 3.5). The effect is amplified at the long BS-RIS and RIS-UE distances considered in rural/macro use cases (e.g., UC-3/6/9/16: up to 2000 m). Because the required RIS aperture grows rapidly with path-loss deficit, a fixed PLE=2 can underestimate the needed element counts by orders of magnitude for these scenarios. The validation at 26 GHz over distances up to 205 m cannot justify extrapolating PLE=2 to long-range rural or FR3 deployments. This is a self-identified limitation, not an external assumption, and it directly affects the quantitative tables that operators would use for sizing decisions.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","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.","tokens_in":8155,"tokens_out":4129,"duration_ms":46213,"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":[{"comment":"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":"Section IV and Section II-C"},{"comment":"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":"Section II-C"},{"comment":"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.","section":"Section II-E and Tables II-IV"}],"minor_comments":[{"comment":"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":"Section II-E"},{"comment":"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":"Section II-B"},{"comment":"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.","section":"Section III and Fig. 4"},{"comment":"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.","section":"Throughout"},{"comment":"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.","section":"Fig. 5"}],"recommendation":"major_revision","confidential_remarks":"The qualitative message is sound and likely useful, but the quantitative message is currently overclaimed relative to the evidence. The fixed PLE=2 assumption and the unspecified averaging over the use-case parameter sets are the two issues that most need work. If the authors add a PLE sensitivity analysis, specify the sampling distributions, and soften the validation claims to match the single-scenario evidence, the paper could become acceptable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","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.\n\nThe 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.\n\nA 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.\n\nThe 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.\n\nThe 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.\n\nI'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.","headline":"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.","tokens_in":8614,"tokens_out":3643,"would_cite":true,"duration_ms":35689,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"RIS panel size must be tuned per deployment, not chosen once for all networks.","keywords":["reconfigurable intelligent surfaces","RIS sizing","path-loss modelling","signal-to-noise ratio","outage probability","sub-6 GHz","FR3","mmWave"],"falsifier":"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.","tokens_in":7633,"feed_emoji":"📡","tokens_out":6878,"duration_ms":60354,"temperature":0.7,"pith_summary":"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.","feed_headline":"RIS sizing is site-specific: table gives minimum panels per scenario","feed_subtitle":"Simulations across sub-6 GHz, FR3, and mmWave show panel size swings with frequency, distance, and environment.","key_machinery":"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.","core_discovery":"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.","pith_inferences":["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."],"forward_implications":["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."],"supporting_citations":[{"why":"Supplies path-loss modeling for RIS-enabled channels that motivates the propagation assumptions in the use cases.","marker":"[3]"},{"why":"Provides indoor and outdoor physical channel modeling for RIS in mmWave bands that informs the angular and distance parameter ranges.","marker":"[4]"},{"why":"Gives path-loss modeling with experimental measurement that grounds the validation approach and propagation expectations.","marker":"[5]"},{"why":"Introduces the scattering-based RIS signal model that the entire simulation uses.","marker":"[6]"},{"why":"Provides the outdoor trial KPI records used to validate the simulator's average SNR behavior.","marker":"[7]"},{"why":"Gives the empirical mmWave path-loss exponent (1.785) used as a benchmark in validation.","marker":"[8]"},{"why":"Supplies standard channel-model parameters for 0.5 to 100 GHz that shape the outdoor use case definitions.","marker":"[9]"},{"why":"Supplies standard deployment geometry and parameter guidance that inform the use case tables.","marker":"[10]"}],"fun_headline_variants":["RIS size? Simulations show it's scenario-dependent","No universal RIS panel size, simulation study finds","Simulations show RIS size needs vary by deployment scenario","RIS sizing: one size doesn't fit all, simulations say"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"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.","fun_headline_variants_meta":{"raw":{"variants":["RIS size? Simulations show it's scenario-dependent","No universal RIS panel size, simulation study finds","Simulations show RIS size needs vary by deployment scenario","RIS sizing: one size doesn't fit all, simulations say"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001072,"raw_usage":{"total_tokens":4470,"prompt_tokens":904,"completion_tokens":3566,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":520,"completion_tokens_details":{"reasoning_tokens":3503}},"tokens_in":520,"tokens_out":3566,"duration_ms":26449,"temperature":1.0,"reasoning_tokens":3503,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:55:22.688913+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"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.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies path-loss modeling for RIS-enabled channels that motivates the propagation assumptions in the use cases."},{"cited_title":"Başar, İ","cited_arxiv_id":null,"evidence_quote":"Provides indoor and outdoor physical channel modeling for RIS in mmWave bands that informs the angular and distance parameter ranges."},{"cited_title":"Tang et al., ``Wireless communications with reconfigurable intelligent surface: Path loss modeling and experimental measurement,\" IEEE Trans","cited_arxiv_id":null,"evidence_quote":"Gives path-loss modeling with experimental measurement that grounds the validation approach and propagation expectations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the scattering-based RIS signal model that the entire simulation uses."},{"cited_title":"Rains, J","cited_arxiv_id":null,"evidence_quote":"Provides the outdoor trial KPI records used to validate the simulator's average SNR behavior."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the empirical mmWave path-loss exponent (1.785) used as a benchmark in validation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies standard channel-model parameters for 0.5 to 100 GHz that shape the outdoor use case definitions."},{"cited_title":"[Online]","cited_arxiv_id":null,"evidence_quote":"Supplies standard deployment geometry and parameter guidance that inform the use case tables."}],"review_version":1}