{"id":"f8731516-6b05-4062-b8fb-5b96a885c972","arxiv_id":"2608.10136","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A continuously moving pinching antenna serving users along its path, antenna roaming, can outperform stop-and-place operation by saving repositioning time at the cost of spatially averaged channel quality.","lead":"This paper studies pinching antennas, small radiating elements that slide along a waveguide, and accounts for the finite time needed to reposition them between users. It compares the standard stop-and-transmit mode with a new roaming mode where the antenna keeps moving while serving users, and shows when roaming wins.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Placement is restricted to equal T/N slots while roaming optimizes interval lengths, so reported roaming gains may conflate continuous motion with unequal time allocation; the fair-comparison claim needs a symmetric benchmark.","rationale":"The reader's weakest assumption focuses on the free-space channel model and constant-speed motion. The channel concern is less damaging than stated: if waveguide attenuation is modeled as a common position-dependent factor A(x) in |h_n(x)|^2, then the inequality r_m(x) >= r_n(x) reduces to A(x)/d_m^2 >= A(x)/d_n^2, so A(x) cancels and Lemma 1's single-crossing property survives. Reflections or small-scale fading would break single crossing, but the paper is explicitly LoS/free-space and does not claim otherwise. The constant-speed assumption is a real simplification, but it is explicit in Eq. (10) and is a modeling choice rather than a hidden comparison bias. The allocation asymmetry, by contrast, is built into the problem formulations themselves and directly feeds the headline numerical claim. Proposition 1 is algebraically correct, but its 'effective-transmission-time gain' term mixes two effects: removal of dedicated positioning time and unequal time allocation across users. The paper never isolates these effects. Fig. 5 is the clearest example: roaming beating ideal placement, which has no positioning overhead, cannot be attributed to avoiding overhead; the text itself attributes it to optimized interval lengths. A reviewer should therefore condition acceptance on a symmetric allocation test or on a clear statement that the comparison is placement-with-equal-slots versus roaming-with-optimized-durations, with correspondingly qualified conclusions. Because the reader already issued CONDITIONAL, my recommended verdict is UNCHANGED; the additional condition is a sharper statement of what must be verified.","tokens_in":18752,"tokens_out":15806,"duration_ms":164216,"concrete_test":"Re-run the Fig. 5 (and Fig. 3) comparisons with a symmetric placement benchmark that optimizes per-user slot durations tau_n subject to sum_{n=1}^N tau_n = T and tau_n >= (x_pt_n - x_pt_{n-1})/v_max, alongside the existing QoS constraints, and compare against the paper's DP roaming. If placement with flexible slots matches or exceeds roaming in the region-width regime where roaming beats ideal placement, the claimed advantage is due to unequal time allocation rather than continuous motion; if roaming still wins, the paper's interpretation survives.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing assumption is not the LoS channel model but the asymmetric allocation of transmission time. In Eq. (9), antenna placement fixes each user's slot to T/N: R_AP_n = (1/N - (x_pt_n - x_pt_{n-1})/(T v_max)) r_n(x_pt_n). In Eqs. (14)-(16), antenna roaming optimizes the service-interval lengths, so R_AR_n = ((x_bd_{n+1}-x_bd_n)/D_x) bar_r_n and the durations are decision variables. The paper's own Remark after Eq. (14) concedes that roaming 'also optimizes user service durations' and has an 'additional degree of freedom.' Consequently, the first term in Proposition 1 measures not only the removal of dedicated repositioning time but also the freedom to shift time toward high-rate users. Fig. 5 makes this explicit: roaming is reported to exceed ideal infinite-speed placement when the region width exceeds about 25 m, and the text attributes this to optimized service-interval lengths while ideal placement 'retains equal service durations.' Since ideal placement has zero positioning overhead, any roaming gain over it cannot be evidence for the paper's stated mechanism of avoiding overhead. Without a symmetric benchmark, the central claim that roaming outperforms placement is not established as a property of continuous motion; it may be an artifact of unequal resource allocation.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper considers a downlink pinching-antenna system with a single antenna moving along a waveguide and serving N users by TDMA, under a finite maximum antenna speed. It contrasts two operation modes: antenna placement, where the antenna stops at optimized service points and each user receives an equal time slot of T/N that includes repositioning time, and antenna roaming, where the antenna moves continuously at constant speed and the trajectory is partitioned into consecutive service intervals whose lengths are optimized. The paper derives an algebraic decomposition of the sum-rate difference between the two modes (Proposition 1), a local quadratic approximation (Proposition 2), a single-crossing property for the instantaneous rate functions (Lemma 1), and an optimal unconstrained service-interval partition (Proposition 3). It then formulates QoS-constrained sum-rate maximization problems for both modes and solves discretized versions via dynamic programming (Sections IV and V). Simulations compare the proposed schemes against fixed-antenna, ideal infinite-speed placement, and uniform schedules, and report that antenna roaming outperforms antenna placement, especially when repositioning overhead is significant.","tokens_in":18972,"tokens_out":3980,"duration_ms":40755,"significance":"If the comparison is made fair, the paper would provide a useful framework for finite-speed movable antenna systems. The proofs of Proposition 1, Lemma 1, and Proposition 3 are correct under the stated channel model, and the DP recursions are standard and globally optimal for the discretized problems. The paper ships no code or data, but the derivations are transparent and the structural result in Lemma 1 is a clean contribution. However, the central performance claim currently conflates two distinct effects: the removal of dedicated positioning overhead and the additional freedom to allocate unequal service durations. Because the manuscript itself acknowledges this extra degree of freedom and even attributes the advantage over ideal placement to it, a major revision with symmetric benchmarks is needed before the main claim is established.","major_comments":[{"comment":"The comparison between antenna placement and antenna roaming is asymmetric in resource allocation. In Eq. (9), placement fixes each user's slot to T/N, while in Eqs. (14)-(16) roaming treats the service-interval lengths (and hence the service durations) as decision variables. Consequently, the first term in Proposition 1 measures not only the elimination of dedicated repositioning time but also the benefit of freely shifting transmission time toward high-rate users. The Remark after Eq. (14) concedes that roaming \"also optimizes user service durations\" and has an \"additional degree of freedom.\" The text accompanying Fig. 5 explicitly explains roaming's ability to outperform ideal infinite-speed placement by optimizing service-interval lengths while ideal placement \"retains equal service durations.\" Since ideal placement has zero positioning overhead, any gain over it cannot be attributed to the removal of repositioning overhead. The paper should add symmetric benchmarks: e.g., placement with optimized unequal slot durations, or roaming with equal-length service intervals, and then compare the corresponding modes under the same resource-allocation flexibility.","section":"Section II-C/D, Eqs. (9), (14), (17), Fig. 5"},{"comment":"The single-crossing property in Lemma 1 and the resulting optimal partition structure in Proposition 3 depend on the specific channel power form |h_n(x)|^2 = \\eta^2/((x-x_n)^2 + y_n^2 + d^2), which includes free-space path loss only. The channel model in Eq. (2) omits in-waveguide attenuation, even though the introduction cites related work (e.g., [16]) that explicitly accounts for the tradeoff between in-waveguide power loss and free-space path loss. With a realistic distance-dependent waveguide attenuation factor, the pairwise instantaneous rate functions need not intersect at most once, and the structural characterization of the optimal service-interval partition may fail. The authors should either include a waveguide-loss term in the channel model or state the conditions under which the single-crossing property and Proposition 3 remain valid in the presence of such loss.","section":"Section II-A, Eq. (2), Lemma 1, Proposition 3"},{"comment":"The exact time-to-space transformation under antenna roaming relies on strictly constant-speed motion, vpin = Dx/T, with no acceleration, deceleration, or speed variation. The paper acknowledges that speed variations are neglected in the placement model (footnote 2), but it does not flag the same idealization for the roaming model, where the assumption is load-bearing for the equivalence between R_AR_n and the spatial integral in Eq. (14). If real actuators require acceleration/deceleration, the position-time mapping becomes nonlinear, the normalized rate expressions change, and the local quadratic interpretation in Proposition 2 would need modification. The paper should state this limitation explicitly and provide a quantitative sensitivity check, for example by simulating a trapezoidal velocity profile.","section":"Section II-C, Eqs. (10)-(14)"}],"minor_comments":[{"comment":"The displayed equations in Remark 5 contain stray text \"bracehtipupleft/bracehtipdownright\" and similar artifacts that appear to be leftover LaTeX macro text; these should be cleaned before publication.","section":"Section II-D, Remark 5"},{"comment":"The symbol N is used both for the set of users and the number of users (e.g., \"N = {1, 2, ..., N}\"). This overload is confusing; use a calligraphic or other distinct symbol for the user set.","section":"Throughout"},{"comment":"The stage-reward function is denoted Fn in both Algorithm 1 (service-point deployment) and Algorithm 2 (service-interval partitioning), even though the two functions have different definitions. Renaming one of them (e.g., G_n for roaming) would improve readability.","section":"Sections IV and V"},{"comment":"The statement \"Under uniform spatial allocation, antenna roaming still achieves a higher sum rate than antenna placement\" is a useful symmetric comparison, but it appears only as a sentence without a dedicated figure; a quantitative comparison of uniform placement versus uniform roaming under identical equal-duration constraints would directly address the fairness concern raised above.","section":"Section VI, Fig. 6"}],"recommendation":"major_revision","confidential_remarks":"The main issue for the editor is the fairness of the central comparison: the paper's headline result that roaming beats placement is not isolated from the unequal-time-allocation effect, and the manuscript itself acknowledges this in the Remark after Eq. (14). The theoretical apparatus is otherwise sound, but the paper should be revised to include symmetric benchmarks or reframed to claim only that roaming offers an additional degree of freedom. The waveguide-attenuation omission is a secondary but real correctness risk for the structural results."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a real contribution, and the math is mostly solid, but the paper's headline comparison is not as fair as it claims. The finite-speed placement model and the roaming mode are genuinely new in the pinching-antenna line, and the algebraic decomposition in Proposition 1 is exact. But roaming is allowed to optimize user service durations while placement is fixed to equal T/N slots. The authors admit the extra degree of freedom in the remark after Eq. (14), then the abstract and conclusions drop that caveat and say roaming consistently outperforms placement. Fig. 5 makes the problem concrete: roaming beats ideal infinite-speed placement, which has zero repositioning overhead, so that gain cannot come from the stated overhead-avoidance mechanism. It comes from shifting transmission time toward high-rate users. A symmetric benchmark—placement with optimized unequal slot lengths, or roaming restricted to equal-length intervals—is needed before the headline claim is credible.\n\nWhat the paper does well: the finite-speed placement model is a sensible generalization of prior instantaneous-placement work; Eq. (17) is a clean, parameter-free decomposition; Lemma 1's single-crossing proof is correct under the LoS channel model; and the DP recursions are standard and globally optimal for the discretized problems. The uniform-roaming results in Figs. 3, 6, and 7 provide some evidence that continuous transmission itself helps, so the core idea probably survives in weakened form.\n\nSoft spots, in proportion: the LoS-only channel model underpins the single-crossing structure and the optimal partition result; if small-scale fading or reflections are significant, the interval structure may fail. The constant-speed, no-acceleration motion is another idealization, and the paper does not flag either assumption as a limitation. The simulations have no error bars or trial counts, and no code or data is provided, so the quantitative claims are not independently reproducible. None of this is fatal by itself, but together it means the present version overstates the case.\n\nThis paper is for researchers working on pinching antennas, movable antennas, or trajectory-aware resource allocation. It deserves a serious referee—the concept is useful and the derivations are careful—but I would send it back for major revision on comparison fairness and simulation reporting.","headline":"Useful, mostly sound pinching-antenna paper whose headline roaming-vs-placement claim is skewed by unequal time allocation; needs a symmetric benchmark before the central comparison is credible.","tokens_in":19517,"tokens_out":3322,"would_cite":true,"duration_ms":37899,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A pinching antenna that serves users while continuously moving can outperform conventional stop-and-serve placement under finite movement speed.","keywords":["pinching-antenna systems","antenna roaming","antenna placement","finite antenna movement speed","sum rate maximization","dynamic programming","service-interval partition","single-crossing property"],"falsifier":"Measure the instantaneous rate functions of two users along a real dielectric waveguide with the antenna moving at constant speed. If the two curves cross more than once (e.g., under multipath or in-waveguide attenuation), Lemma 1's single-crossing property is false and the predicted optimal partition boundaries will not coincide with rate intersections. Alternatively, simulate the same system with realistic acceleration and deceleration phases: if the sum-rate gap between roaming and placement deviates from Proposition 1's constant-speed prediction, the exact decomposition's dependence on the speed model is exposed.","tokens_in":18524,"feed_emoji":"📡","tokens_out":7412,"duration_ms":62105,"temperature":0.7,"pith_summary":"This paper asks what happens to a pinching-antenna system—a radiating point formed by attaching a dielectric particle to a waveguide—when repositioning the antenna takes real time. Conventional antenna placement assumes the antenna can jump instantly between optimized service points; under a finite movement speed, every stop eats into the transmission slot. The paper proposes antenna roaming, in which the antenna glides continuously along the waveguide and serves each user over a spatial interval instead of at a point, and it builds a common cycle-duration model to compare the two fairly. The central result is an exact decomposition of the sum-rate difference between the two modes into a gain from transmission time and a loss from spatially averaged channel quality. A sympathetic reader cares because the decomposition turns a hardware timing nuisance into a design variable: roaming avoids positioning overhead and, in the paper's simulations, approaches or even beats ideal infinite-speed placement when repositioning overhead is large or channel-rate curves are flat.","feed_headline":"Roaming antenna beats stop-and-serve at finite speed","feed_subtitle":"Serving while moving keeps the link busy and can match ideal placement when repositioning is slow.","key_machinery":"The central object is the unified cycle-duration comparison embodied in Proposition 1's decomposition (Eq. (17)), which rewrites the sum-rate difference between roaming and placement so that the first term captures uninterrupted transmission time and the second captures the channel-quality loss from spatial averaging. The supporting mechanism is the free-space channel power gain $|h_n(x)|^2 = \\eta^2/((x-x_n)^2+y_n^2+d^2)$, which makes each user's rate function single-peaked at the user's projection $x_n$ and gives any two users at most one rate intersection (Lemma 1). That single-crossing property is what lets the paper describe the unconstrained optimal roaming partition in closed structural form: assign each antenna position to the user with the highest instantaneous rate, with boundaries at rate-function intersections.","core_discovery":"The paper's central claim is that, when the pinching antenna moves at a finite speed, operating it in a roaming mode—serving users continuously as it slides along the waveguide—can outperform the conventional stop-at-a-service-point placement mode, because the time lost to repositioning outweighs the loss from averaging the channel over an interval. The formal engine is an exact identity (Proposition 1, Eq. (17)) splitting the sum-rate difference between the two modes into an effective-transmission-time term and a spatially averaged channel-quality term; under a second-order approximation near each user's projection, the channel term becomes a function of squared distance from the projection and of interval length (Proposition 2). For the unconstrained roaming problem, the paper proves a single-crossing property of the instantaneous rate functions (Lemma 1) and characterizes the optimal service-interval partition as assigning every antenna position to the currently strongest user, with boundaries at rate-function intersections (Proposition 3). Both rate-maximization problems are solved by dynamic programming over discretized positions, and simulations show that DP-optimized roaming approaches, and for wide service regions exceeds, the sum rate of ideal infinite-speed placement.","pith_inferences":["Editorial inference: the same stop-and-serve versus move-and-serve tradeoff should arise in other reconfigurable-antenna architectures whose repositioning is finite-speed, so Proposition 1's decomposition is a candidate general design principle beyond pinching antennas.","Editorial inference: the single-crossing structure means the unconstrained roaming partition is a kind of weighted Voronoi partition in the coordinate $x$; for the free-space model, boundaries solve a quadratic equation in $x$, so the optimal intervals can be computed in closed form rather than by DP, at least when QoS constraints are inactive.","Testable extension: if movement energy or mechanical wear is counted, roaming's continuous actuation may offset its time gain; a hybrid mode that roams through low-rate regions and pauses at high-rate points could dominate both pure modes, a design the paper does not explore.","Testable extension: the paper's constant-speed, no-acceleration motion is optimistic; one could re-run the comparison with trapezoidal velocity profiles to see whether the roaming advantage survives finite acceleration, which would tell whether the qualitative conclusion depends on the exact speed model."],"forward_implications":["Roaming is a viable alternative whenever repositioning overhead is non-negligible: DP-based roaming approaches ideal infinite-speed placement and, for wide service regions, exceeds it (Figs. 3 and 5).","The optimal unconstrained roaming partition has a simple rule: serve, at each antenna position, the user with the highest instantaneous rate; boundaries between users sit at intersections of their rate curves (Proposition 3).","Longer waveguides favor roaming because placement pays both repositioning time and degrading channel quality, whereas roaming only pays channel degradation (Fig. 4).","Because roaming's average rate does not depend on cycle duration $T$ (the spatial normalization $1/D_x$ cancels the temporal one), its performance is cycle-duration independent, unlike finite-speed placement whose rate grows with $T$ (Fig. 6).","The DP algorithms have $O(NK^2)$ worst-case complexity over $K$ candidate positions and are globally optimal for the discretized problems, so the tradeoff can be optimized in practice."],"supporting_citations":[{"why":"Introduces the pinching-antenna concept (dielectric particles on waveguides) that defines the system architecture.","marker":"[7]"},{"why":"Establishes the flexible-antenna/pinching-antenna perspective and the base channel and rate setup this paper extends.","marker":"[8]"},{"why":"Provides the continuous-placement model with waveguide/free-space path-loss tradeoff that the finite-speed placement model generalizes.","marker":"[16]"},{"why":"Documents movement energy/cost in pinching-antenna systems, motivating the repositioning-delay modeling.","marker":"[26]"},{"why":"Analyzes motion power consumption in pinching-antenna systems, another finite-speed constraint this paper abstracts as time overhead.","marker":"[27]"},{"why":"Supplies the discrete-trajectory communication model whose limiting form supports the spatial-average roaming rate expression in Eq. (14).","marker":"[31]"},{"why":"Provides the dynamic programming principle used to solve the discretized sequential decision problems.","marker":"[34]"}],"fun_headline_variants":["Roam to serve: moving antenna beats stop-and-place","When repositioning costs time, roaming antenna wins","Slide and serve: roaming antenna beats fixed placement","Antenna roaming: serving on the move beats stopping"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the pure free-space, constant-speed channel model: each user's rate curve is single-peaked and any two curves cross at most once, and antenna position maps linearly to time, so the optimal-interval structure and exact gain decomposition depend on this idealized geometry; if in-waveguide loss, reflections, fading, or acceleration break these, the derived structure can fail.","fun_headline_variants_meta":{"raw":{"variants":["Roam to serve: moving antenna beats stop-and-place","When repositioning costs time, roaming antenna wins","Slide and serve: roaming antenna beats fixed placement","Antenna roaming: serving on the move beats stopping"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000196,"raw_usage":{"total_tokens":1384,"prompt_tokens":991,"completion_tokens":393,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":607,"completion_tokens_details":{"reasoning_tokens":331}},"tokens_in":607,"tokens_out":393,"duration_ms":4075,"temperature":1.0,"reasoning_tokens":331,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:11:29.858770+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the instantaneous rate functions of two users along a real dielectric waveguide with the antenna moving at constant speed. If the two curves cross more than once (e.g., under multipath or in-waveguide attenuation), Lemma 1's single-crossing property is false and the predicted optimal partition boundaries will not coincide with rate intersections. Alternatively, simulate the same system with realistic acceleration and deceleration phases: if the sum-rate gap between roaming and placement deviates from Proposition 1's constant-speed prediction, the exact decomposition's dependence on the speed model is exposed.","supporting_citations":[{"cited_title":"Pin ching- antenna systems: Waveguide-power loss and free-space path loss trade- off,","cited_arxiv_id":null,"evidence_quote":"Provides the continuous-placement model with waveguide/free-space path-loss tradeoff that the finite-speed placement model generalizes."},{"cited_title":"Energy-efficient de sign for pinching-antenna system with movement energy consumption ,","cited_arxiv_id":null,"evidence_quote":"Documents movement energy/cost in pinching-antenna systems, motivating the repositioning-delay modeling."},{"cited_title":"Join t radiation power, antenna position, and beamforming optimization for pinching- antenna systems with motion power consumption,","cited_arxiv_id":null,"evidence_quote":"Analyzes motion power consumption in pinching-antenna systems, another finite-speed constraint this paper abstracts as time overhead."},{"cited_title":"Joint trajectory and commu nication design for multi-UA V enabled wireless networks,","cited_arxiv_id":null,"evidence_quote":"Supplies the discrete-trajectory communication model whose limiting form supports the spatial-average roaming rate expression in Eq. (14)."}],"review_version":1}