{"id":"8ca5b594-707e-4e0b-bc26-8fade2729ea2","arxiv_id":"2509.04885","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"For pinching-antenna IoT systems in circular rooms, this paper derives outage and rate expressions and shows that with partial waveguide coverage and propagation loss, performance is non-monotonic in waveguide length.","lead":"Pinching-antenna systems use movable antennas along a waveguide to serve IoT devices; this paper analyzes them inside circular rooms and derives formulas for outage and data rate. It finds that when the waveguide is shorter than the room diameter and suffers signal loss, there is an optimal waveguide length that balances coverage against attenuation.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The non-monotonic waveguide-length trade-off is likely an artifact of the closest-point placement rule: with loss-aware PA placement, per-user SNR is non-decreasing in l, so outage/rate become monotonic.","rationale":"The reader's weakest_assumption identifies Eq. (26) as the soft spot, and my stress-test sharpens that concern from 'could change' to 'provably changes the qualitative result.' Under any placement rule that actually optimizes SNR over the available waveguide interval [-l,l], the feasible set is nested in l, so the best achievable SNR for each user is non-decreasing in l. The advertised non-monotonic trade-off and the claimed optimal-length shift with alpha are therefore properties of the closest-point rule, not of PASS itself. This is the single most load-bearing concern because it attacks the paper's headline contribution rather than its peripheral formulas. I nonetheless keep the reader's CONDITIONAL verdict rather than moving to REJECT: the derivations and Monte Carlo validations may be internally correct for the stated placement rule, and the paper could become acceptable by reframing the contribution as an analysis of the closest-point strategy, explicitly stating that the design trade-off is specific to that rule, and adding a sensitivity study of loss-aware placement. The reader's other concerns (closed-form overclaim, Lemma 6 inconsistencies, unstated quadrature order) remain valid but are secondary; they are fixable without changing the model, whereas the placement-rule issue determines whether the central conclusion survives. The proposed numerical test would settle the matter directly by reusing the paper's own simulation setup.","tokens_in":18568,"tokens_out":8304,"duration_ms":78826,"concrete_test":"Numerically recompute outage probability and average achievable rate versus l for the PWL scenario, replacing Eq. (26) by x_p* = argmax_{x_p in [-l,l]} [ -alpha(x_p+l) - log(h^2 + y_u^2 + (x_u - x_p)^2) ] per Monte Carlo user, using the same parameters as Figs. 4 and 7 (r = 25 m, h = 10 m, gamma_th = 100, alpha in {0.01, 0.02, 0.03, 0.04}, same transmit SNR). A simple implementation is to evaluate the SNR on a fine grid of x_p for each user draw, or solve the cubic stationarity condition. If outage is non-increasing and rate is non-decreasing in l, the non-monotonic optimum is an artifact of Eq. (26) and the design guideline must be revised. This test is cheap and uses the same simulation framework already in the paper.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline design trade-off (Section IV.B, Figs. 4 and 7) is established under the placement rule Eq. (26), x_p = clamp(x_u, -l, l), which minimizes free-space distance but ignores the attenuation factor e^{-alpha(x_p+l)} entering the SNR in Eq. (36). This rule is not merely suboptimal; it is the mechanism producing the claimed non-monotonicity. For any fixed user (x_u, y_u), the SNR under a loss-aware rule is gamma*(l) = max_{x_p in [-l,l]} eta Pt e^{-alpha(x_p+l)} / [sigma^2 (h^2 + y_u^2 + (x_u - x_p)^2)]. Since the feasible interval [-l,l] is nested in l, gamma*(l) is non-decreasing in l for every user; outage probability is non-increasing and average achievable rate is non-decreasing in l. Thus no interior optimum in l can exist under optimal placement, and the paper's 'optimal length decreases with alpha' conclusion depends wholly on forcing x_p to the nearest coordinate/endpoint. The derivative condition for the unconstrained optimum is -alpha - 2(x_p - x_u)/(h^2 + y_u^2 + (x_u - x_p)^2) = 0, giving x_p* < x_u for alpha > 0; the closest-point rule is optimal only at alpha = 0. The formulas and Monte Carlo plots may be correct for the stated rule, but they do not support the general practical guideline claimed in the abstract unless the paper explicitly restricts the conclusion to this fixed suboptimal rule. Secondary but related: Lemmas 2, 3, and 6 are quadrature sums, not closed forms, and Lemma 6's weighting and pi factors are inconsistent with Lemma 3, so the rate verification is also shaky.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper analyzes downlink pinching-antenna systems (PASS) deployed in a circular indoor environment, covering full-coverage and partial-coverage waveguides, each with and without exponential waveguide attenuation. The pinching antenna is placed at the point on the waveguide closest to the device (Eq. (26)), and the device location is uniformly distributed over a disk. For the four configurations, the authors derive expressions for outage probability and average achievable rate, using Gauss-Chebyshev quadrature for the average rate in the lossy cases, and validate the expressions by Monte Carlo simulation. The headline finding is that, for partial coverage with propagation loss, outage probability and average achievable rate are non-monotonic in the waveguide half-length l, and the optimal l decreases as the attenuation coefficient alpha increases (Figs. 4 and 7).","tokens_in":1672,"tokens_out":1948,"duration_ms":80213,"significance":"If the results are correct for the stated model, the paper provides a useful extension of PASS performance analysis to non-rectangular geometries and partial-coverage waveguides, and the outage-probability derivations are genuinely analytical. The Monte Carlo validation and the systematic treatment of four scenarios are strengths. However, the headline non-monotonic design trade-off is obtained under a fixed suboptimal placement rule that ignores waveguide attenuation, and it disappears when the antenna is placed to maximize the actual SNR. In addition, the abstract's promise of closed-form rate expressions is not met: Lemmas 2, 3, and 6 are quadrature approximations, and Lemma 6 as printed is not a valid Gauss-Chebyshev approximation. These issues materially affect the paper's central claims and practical guidelines.","major_comments":[{"comment":"The non-monotonic dependence on the waveguide length l is established only under the placement rule x_p = clamp(x_u, -l, l), which minimizes the free-space distance but ignores the waveguide attenuation factor e^{-alpha(x_p+l)} that appears in the SNR (Eq. (36)). For any fixed user, if the PA is placed to maximize the SNR, the optimized SNR gamma*(l) = max_{x_p in [-l,l]} eta P_t e^{-alpha(x_p+l)} / [sigma^2 (h^2 + y_u^2 + (x_u - x_p)^2)] is non-decreasing in l because the feasible interval [-l,l] is nested as l grows. Consequently the outage probability is non-increasing and the average achievable rate is non-decreasing in l, so the interior optimum reported in Figs. 4 and 7 cannot exist under a loss-aware placement rule. The closest-point rule is optimal only in the alpha = 0 limit. The abstract's conclusion that the optimal length decreases as the attenuation coefficient increases is therefore an artifact of a fixed suboptimal placement heuristic, not a property of PASS with rational antenna placement. The manuscript should explicitly restrict the headline claim to the fixed rule in Eq. (26) and ideally compare it with a loss-aware placement rule; as written, the practical guidelines are overstated. A related consequence visible in Figs. 3 and 6 is that partial coverage is reported to outperform full coverage for alpha = 0.02 and 0.04; under loss-aware placement, full coverage always dominates partial coverage because its feasible placement interval is larger.","section":"Section IV.B, Eq. (26), and Figs. 4 and 7"},{"comment":"The abstract and Section I state that closed-form expressions for the average achievable rate are derived for all four scenarios. However, Lemmas 2, 3, and 6 give finite Gauss-Chebyshev sums whose accuracy depends on the number of quadrature points (V or n); these are numerical quadrature approximations, not closed-form expressions. The claims should be reworded to say 'analytical approximations' or 'semi-analytical expressions', and each lemma should state explicitly that the displayed formula is a quadrature approximation rather than an exact closed form.","section":"Abstract and Lemmas 2, 3, and 6"},{"comment":"Equation (53) is not the Gauss-Chebyshev quadrature approximation claimed in Appendix D. A standard Gauss-Chebyshev approximation of the integral from -l to l of F(x) dx is (pi l / n) sum_{k=1}^n sqrt(1 - t_k^2) F(l t_k), but Eq. (53) omits both the factor pi and the Chebyshev weights sqrt(1 - t_k^2), and it also changes the factor in front of the F1 term to 2l instead of l. The same inconsistency affects the F2 and F3 terms. As printed, Eq. (53) cannot reproduce the Monte Carlo results or the plotted PWL rate curves. Please correct the quadrature formula (or redefine F1-F3 so that the printed expression is correct) and verify the numerical implementation against the corrected formula.","section":"Lemma 6, Eq. (53), and Appendix D"}],"minor_comments":[{"comment":"The feed point is defined as psi_a = (0,0,h), but Eq. (5) states |psi_p - psi_a| = x_p + r, which corresponds to a feed at (-r,0,h). These definitions are inconsistent for negative x_p, and the discrepancy affects the SNR expression. Please clarify the actual feed-point location for both full- and partial-coverage waveguides.","section":"Section II, Eqs. (4) and (5)"},{"comment":"Equation (32) is in the subsection without propagation loss, yet it contains the factor e^{-alpha(x_u+r)}. The exponential attenuation factor should be removed from this equation to match the lossless SNR in Eq. (28) and the subsequent use of A = eta P_t / (sigma^2 gamma_th) - h^2.","section":"Section IV.A, Eq. (32)"},{"comment":"In the second line of Eq. (B.2), the two dilogarithm terms are identical: (1/alpha) Li2(-eta P_t / Psi(y_u)) appears twice with opposite signs. The first term should involve v1 and the second v2, as in the final expression in Lemma 2. Please correct the derivation.","section":"Appendix B, Eq. (B.2)"},{"comment":"The definitions of v1 and v2 use y_u, but the quadrature nodes in Eq. (25) are denoted y_m. Please make the notation consistent so that the substitution into the dilogarithm terms is unambiguous.","section":"Lemma 2, Eq. (25)"},{"comment":"Reference [29] is listed as 'S. Edition' and should be properly cited as Gradshteyn and Ryzhik, Table of Integrals, Series, and Products, with the edition and publisher information.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The central concern is that the headline non-monotonic trade-off is produced by a suboptimal placement rule and would vanish under a loss-aware placement; this should be addressed by reframing the contribution or by adding an optimal-placement analysis. The quadrature errors in Lemma 6 and the overstatement of 'closed-form' results are fixable but need to be corrected before the paper can be considered for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Hi [Name],\n\nQuick take on arXiv:2509.04885. It is a competent performance analysis of pinching-antenna systems in a circular room, with full vs partial waveguide coverage and with/without waveguide loss. The genuinely new parts are the circular geometry and the partial-coverage regime, not treated in the PASS papers they cite. Monte Carlo matches their formulas for the model as stated.\n\nThe main soft spot, which I think is the thing that matters most, is the load-bearing design claim in Section IV.B. The non-monotonic dependence of outage/rate on waveguide length and the “optimal length decreases with alpha” conclusion are computed under a fixed closest-point placement rule (Eq. 26): the PA sits at the waveguide point horizontally nearest the device. That rule minimizes free-space distance but ignores attenuation cost. For any fixed user, the set of achievable PA positions is monotone increasing in l; if you instead place the PA to maximize the actual SNR, per-user SNR is non-decreasing in l, so outage is non-increasing and rate non-decreasing in l. The interior optimum disappears. The paper never claims the rule is optimal, but the abstract and conclusions present the trade-off as a practical design guideline, and that is misleading without a clear caveat. This is not a mathematical error; it’s a modeling choice that drives the headline result.\n\nSecond-order issues: the abstract promises closed-form expressions, but Lemmas 2, 3, and 6 are Gauss-Chebyshev quadrature approximations. That is fine if labeled, but Lemma 6’s formula is missing the Chebyshev weights and the π/n factor; it is inconsistent with Lemma 3 and with Appendix D. There are also typos in the feed point definition, Eq. (14), and Appendix B, plus no quadrature order or Monte Carlo run counts reported. All fixable, but they need fixing before publication.\n\nThe math appears otherwise sound: Lemma 1 is a genuine closed form, the outage derivations are standard, and the Monte Carlo agreement supports the formulas for the stated rule. The paper is aimed at PASS/IoT performance analysts; it is a useful reference for that community even as a negative example of why placement optimization matters.\n\nI’d send it to review rather than desk-reject — it deserves referee time — but my report would recommend major revision, mainly to reframe the non-monotonic result as an analysis of the closest-point heuristic and to clean up the quadrature claims and formulas.\n\nBest,","headline":"A competent PASS performance analysis in a circular room whose headline non-monotonic waveguide-length trade-off is an artifact of the suboptimal closest-point placement rule; with optimal placement the trade-off disappears.","tokens_in":19469,"tokens_out":5455,"would_cite":false,"duration_ms":48081,"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":"This paper develops closed-form outage and rate expressions for pinching-antenna IoT systems in circular rooms and shows that, with lossy partial-coverage waveguides, performance is non-monotonic in waveguide length, with the optimal…","keywords":["pinching-antenna systems","dielectric waveguides","outage probability","average achievable rate","circular indoor environment","waveguide propagation loss","partial coverage","IoT networks"],"falsifier":"Simulate or measure outage probability and average rate in the stated circular room with a partial waveguide for a fixed attenuation coefficient and a sweep of waveguide lengths from small to the full diameter. If outage keeps decreasing as length approaches the diameter, with no interior minimum for any $\\alpha>0$, the non-monotonic claim fails. Separately, recompute performance under a loss-aware placement rule that balances free-space distance against $e^{-\\alpha(x_p+r)}$; if the interior optimum disappears, the paper's design trade-off is an artifact of the closest-point rule.","tokens_in":18370,"feed_emoji":"📡","tokens_out":5890,"duration_ms":51151,"temperature":0.7,"pith_summary":"This paper tries to establish how pinching-antenna systems behave when the room is circular and the waveguide covers only part of it, with realistic signal loss inside the waveguide. It derives closed-form expressions for outage probability and average achievable rate in four configurations: full or partial waveguide coverage, with or without propagation loss. The central finding is that in the partial-coverage, lossy case the outage probability and rate are non-monotonic in the waveguide length: extending the guide first helps, then hurts, and the optimal length moves downward as the attenuation coefficient increases. These formulas give a quantitative design rule: for a given loss per meter, there is a finite best waveguide length rather than \"the longer the better.\" If true, they turn waveguide sizing into a solvable trade-off between free-space path loss and accumulated guide attenuation.","feed_headline":"IoT outage falls then rises as waveguide length stretches","feed_subtitle":"Closed-form analysis finds a sweet-spot guide length that shrinks as waveguide loss grows.","key_machinery":"The engine of the analysis is the SNR identity $\\gamma = \\frac{\\eta P_t e^{-\\alpha(x_p+r)}}{\\sigma^2(h^2+y_u^2+(x_u-x_p)^2)}$, which couples the two competing effects: free-space path loss shrinks as the antenna gets closer to the device, while the exponential term $e^{-\\alpha(x_p+r)}$ grows with the distance the signal travels along the waveguide from the feed. For partial coverage, the placement rule $x_p = \\mathrm{clamp}(x_u,-l,l)$ makes the distribution of the horizontal distance $D$ piecewise explicit, with a central segment $D=y_u$ and side segments involving $x_u\\pm l$; the cumulative distribution function of $D$ yields outage in closed form, and the same CDF is fed through integral identities and Gauss-Chebyshev quadrature to produce the average-rate expressions. This geometric-propagation coupling is what produces the interior optimum in $l$.","core_discovery":"The paper's central discovery is that, under the stated circular-room model, the received SNR for a pinching antenna at position $x_p$ serving an IoT device at $(x_u,y_u)$ is $\\gamma = \\frac{\\eta P_t e^{-\\alpha(x_p+r)}}{\\sigma^2(h^2+y_u^2+(x_u-x_p)^2)}$, where $\\alpha$ is the waveguide attenuation coefficient, $h$ the antenna height, and $r$ the room radius. Under the closest-point placement rule $x_p = \\mathrm{clamp}(x_u, -l, l)$ for a partial waveguide of half-length $l$, the device-distance distribution and the exponential guide loss combine to make outage and rate first improve and then degrade as $l$ grows beyond an optimum. The paper derives that the optimal $l$ shrinks as $\\alpha$ increases, and it shows that full coverage can underperform partial coverage once attenuation is non-negligible, because the accumulated feed-to-antenna loss outweighs the benefit of always reaching the device. Closed-form expressions are validated against Monte-Carlo simulation.","pith_inferences":["The closest-point placement rule is a modeling choice; a loss-aware rule that trades free-space distance against $e^{-\\alpha(x_p+r)}$ could shift or remove the interior optimum, so the quantitative trade-off is tied to that rule.","The same geometric-propagation coupling likely produces a similar length optimum in other room shapes, making the non-monotonic response a testable prediction for rectangular or elliptical deployment areas.","For multiple IoT devices, averaging these per-device expressions over a non-uniform spatial density would move the optimal length toward device-dense regions, an extension the paper does not develop.","A measurement campaign along a real dielectric waveguide with known $\\alpha$ could check whether the predicted finite optimum appears in practice, which would also test the exponential-attenuation model."],"forward_implications":["With a lossy partial-coverage waveguide, outage probability and average achievable rate are non-monotonic in waveguide length, so installing the longest possible waveguide is not optimal.","The optimal waveguide half-length decreases as the attenuation coefficient $\\alpha$ increases, so higher-loss waveguides should be deployed shorter.","Full coverage can be worse than partial coverage once $\\alpha$ is non-negligible, because accumulated feed-to-antenna loss outweighs the benefit of always reaching the device.","The closed-form expressions let operators evaluate outage and rate for a candidate length directly, replacing Monte-Carlo simulation for design studies."],"supporting_citations":[{"why":"Supplies the pinching-antenna channel model, including free-space path loss, waveguide phase shift, and exponential attenuation, that the paper extends to circular and partial-coverage geometries.","marker":"[25]"},{"why":"Provides the integral identities used to evaluate the closed-form outage and rate expressions in the appendices.","marker":"[29]"},{"why":"Provides the Gauss-Chebyshev quadrature rule used to turn the average-rate integrals into closed-form approximations.","marker":"[30]"}],"fun_headline_variants":["Optimal pinching-antenna guide length shrinks with loss","IoT outage non-monotonic: guide length sweet spot depends on loss","Full-coverage waveguide loses to partial when attenuation high","Pinching antennas: closed-form outage for circular IoT rooms"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the pinching antenna is always placed at the closest point on the waveguide to the device, a rule that ignores the waveguide attenuation cost of the feed-to-antenna distance; if a loss-aware placement rule were used, the non-monotonic optimum and the full-versus-partial coverage comparisons could change.","fun_headline_variants_meta":{"raw":{"variants":["Optimal pinching-antenna guide length shrinks with loss","IoT outage non-monotonic: guide length sweet spot depends on loss","Full-coverage waveguide loses to partial when attenuation high","Pinching antennas: closed-form outage for circular IoT rooms"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000256,"raw_usage":{"total_tokens":1586,"prompt_tokens":965,"completion_tokens":621,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":581,"completion_tokens_details":{"reasoning_tokens":550}},"tokens_in":581,"tokens_out":621,"duration_ms":6328,"temperature":1.0,"reasoning_tokens":550,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:25:33.333222+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate or measure outage probability and average rate in the stated circular room with a partial waveguide for a fixed attenuation coefficient and a sweep of waveguide lengths from small to the full diameter. If outage keeps decreasing as length approaches the diameter, with no interior minimum for any $\\alpha>0$, the non-monotonic claim fails. Separately, recompute performance under a loss-aware placement rule that balances free-space distance against $e^{-\\alpha(x_p+r)}$; if the interior optimum disappears, the paper's design trade-off is an artifact of the closest-point rule.","supporting_citations":[{"cited_title":"Performance analysis of pinching- antenna systems,","cited_arxiv_id":null,"evidence_quote":"Supplies the pinching-antenna channel model, including free-space path loss, waveguide phase shift, and exponential attenuation, that the paper extends to circular and partial-coverage geometries."},{"cited_title":"Table of integrals, series, and products,","cited_arxiv_id":null,"evidence_quote":"Provides the integral identities used to evaluate the closed-form outage and rate expressions in the appendices."},{"cited_title":"On performance of intelligent reflecting surface aided wireless powered IoT network with HIs,","cited_arxiv_id":null,"evidence_quote":"Provides the Gauss-Chebyshev quadrature rule used to turn the average-rate integrals into closed-form approximations."}],"review_version":2}