{"id":"61fd4a95-47a7-4b9c-8acf-0f5c010d9cb8","arxiv_id":"1908.09047","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"For indoor THz links, the optimal antenna beamwidth is set by a tradeoff between gain and mobility-induced misalignment, and this optimum shifts with user mobility type and access point placement.","lead":"This paper uses a simulation of indoor terahertz links to argue that for moving users, the best antenna beamwidth is a tradeoff between signal gain and outages, and that this optimum depends on how the user moves and where the access point sits. Generalists might read it because it gives concrete engineering guidance for future indoor wireless systems, but the guidance rests on a model with several unspecified choices.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Reported δopt values (23°, 25°) exceed S1's individual peak rotation amplitudes (≈15.5°), implying an unstated rotation-combination rule in the outage model.","rationale":"The reader's weakest assumption identified the outage model and the borrowed rotation amplitudes as load-bearing. My stress-test concurs and sharpens the concern: the numerical values in the headline claim are internally inconsistent with the naive reading of the stated outage model. For S1, the largest individual rotation is 15.5°, so any beamwidth greater than that would avoid rotation-induced outage under a per-axis threshold; since smaller beamwidths yield higher gain, the optimum should be just above 15.5°, not 23–25°. The reported values match the vector magnitude of the three rotations, implying an unstated rule for combining yaw, pitch, and roll (or an equally unstated effect of location movement on the angle of arrival). The paper does not provide this rule, an explicit outage probability function, the number of users M, room dimensions, or the beam-steering/scheduling procedure. Therefore, the specific δopt values are not derivable from the manuscript as written. However, the qualitative claim that an optimum exists and shifts with mobility class and AP placement is plausible and would survive a range of reasonable models. The paper is an early simulation study without code or full simulation details; the appropriate outcome is to require clarification and reproducibility, which matches the reader's CONDITIONAL verdict. My analysis does not change that verdict, so I recommend UNCHANGED. A single targeted check—reimplementing the outage rule explicitly—would settle whether the discrepancy arises from a hidden combination rule or from another unspecified simulation element, and would determine whether the reported optima are artifacts.","tokens_in":8878,"tokens_out":7203,"duration_ms":75347,"concrete_test":"Recreate Fig. 6 from the stated text under two explicit outage rules: (i) outage if max(ω_y, ω_p, ω_r) > δ, and (ii) outage if sqrt(ω_y² + ω_p² + ω_r²) > δ. Use Table II peak amplitudes, sinusoidal waveforms with the described noise, the random waypoint model, and M plus room geometry from the authors' code or from natural defaults (e.g., M = 100, 10 m × 10 m × 3 m, AP placements as in Fig. 4). Sweep δ from 1° to 40° in 1° steps and extract the coverage-optimal δ for S1 in Scenarios B and C. If rule (i) gives δopt ≈ 16° instead of 23°/25°, and rule (ii) gives ≈ 25–26°, then the published optima depend on the unstated combination rule and the central claim is not uniquely determined by the paper.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim is that for S1 users the coverage-optimal beamwidths are 23° (Scenario B) and 25° (Scenario C). Section IV states that outages occur when \"the rotation angle of the body (yaw, pitch, and roll) is higher than the beamwidth.\" Table II lists S1 peak individual rotations as ω_y=15.5°, ω_p=13.8°, ω_r=15°. If the outage condition were \"any individual rotation exceeds δ,\" then any δ>15.5° would eliminate rotation-induced outage; since narrower beams give higher gain, the optimum would lie just above 15.5°, not at 23–25°. The reported values instead closely match the Euclidean norm of the three rotations: sqrt(15.5² + 13.8² + 15²) ≈ 25.6°. This indicates the simulation must combine rotations (or include additional angle-of-arrival effects from location movement), but the paper never states this rule. The same pattern appears for S2 fast walk (individual peaks 4°, 5°, 5°; norm ≈ 8.1°) and S3 (5°, 3°, 1°; norm ≈ 5.9°). Because the claimed optima are computed from this unstated combination rule, the headline result is not reproducible from the text alone. This is structural: Fig. 6 and the δopt values are the paper's main contribution, and they rest on an outage model that is underspecified exactly where it matters.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper considers indoor THz communication with a single access point (AP) and studies how system parameters—humidity, distance, frequency windows, antenna beamwidth, AP placement, and user mobility type—affect throughput and user coverage. It first develops a static model that selects frequency windows based on distance and water-vapor concentration, and derives a closed-form minimum beamwidth for a target data rate (Eq. 2). It then introduces a mobility model characterized by a six-dimensional instability parameter Θ = [Δx, Δy, Δz, Δω_y, Δω_p, Δω_r], classifies users into three service types (S1 high mobility, S2 constrained mobility with fast/slow walking, S3 low mobility), and simulates a random-waypoint indoor scenario with three AP placements to evaluate peak throughput and average user coverage as a function of beamwidth. The central claim is that for mobile users there exist mobility- and placement-dependent optimal beamwidths—for example, δopt = 23° for S1 in Scenario B and δopt = 25° in Scenario C—which resolve the 'beamwidth dilemma' between antenna gain and rotation-induced outage.","tokens_in":9167,"tokens_out":6129,"duration_ms":59952,"significance":"If the quantitative results are reliable, the paper provides a useful parameter-based approach for adaptive beamwidth selection in indoor THz networks, addressing a real deployment concern: narrow beams increase gain but are sensitive to small-scale body/device motion. The static model with Eq. (2) gives a simple closed-form bound, and the instability-parameter framework is a reasonable way to organize mobility scenarios. The qualitative finding that the optimal beamwidth depends on mobility class and AP placement is plausible and consistent with prior work on small-scale mobility in THz/mmWave bands. However, the paper's main quantitative contribution rests on a simulator whose outage model and instance parameters are not fully specified, and the reported optima appear to imply an unstated combination rule for the rotation components; this limits the reproducibility and transferability of the specific δopt values until the missing details are supplied.","major_comments":[{"comment":"The outage model that maps the six-dimensional instability parameter Θ to a beamwidth-dependent outage is never written down. The text in Section IV states that outages occur when 'the rotation angle of the body (yaw, pitch, and roll) is higher than the beamwidth', but it does not specify how the three rotation components are combined into a single angle. This is load-bearing because Table II lists S1 peak rotations of ω_y = 15.5°, ω_p = 13.8°, and ω_r = 15°; under a max-component rule the coverage-optimal beamwidth for S1 would be just above 15.5°, yet the paper reports δopt = 23° (Scenario B) and 25° (Scenario C), values close to the Euclidean norm of the three rotations (≈25.6°). The apparent use of an unstated norm-like combination rule makes the headline numerical results non-reproducible from the text alone.","section":"Section IV, Fig. 6 and Table II"},{"comment":"The simulation instance is underspecified. The number of active users M, room dimensions, simulation duration, random-waypoint parameters, the noise applied to the rotational movements, and the threshold used for 'clipping off the data rate dips' in the frequency-window selection are not reported. Without these values, the coverage curves in Fig. 6 and the quantitative ranking of Scenarios A, B, and C cannot be audited, reproduced, or transferred to a different room geometry.","section":"Section IV, Figs. 5 and 6"},{"comment":"The mobility amplitudes assigned to the three service types are taken from unrelated motion-capture studies (walking [15,16], running [17], VR tracking [18]) and mapped to S1/S2/S3 by hand, and the sinusoidal rotation pattern is described only verbally. Because the optimal beamwidth is largely dictated by these amplitude choices, the paper should either justify their representativeness for the target THz device placements (head, hand, body) or provide a sensitivity analysis showing how δopt changes with the amplitude assumptions; otherwise the specific optimal values are an artifact of the chosen table.","section":"Table II and Sections III–IV"},{"comment":"The paper should explicitly acknowledge that the existence of an optimal beamwidth is a direct consequence of the assumed outage model: if outages occur whenever the rotation angle exceeds the beamwidth, then any δ below the characteristic rotation amplitude yields high outage probability and any δ above it eliminates rotation-induced outage, so the existence of a peak is built into the simulation. The contribution is therefore the parameterization and the specific tradeoff curves, not the qualitative existence of an optimum; this distinction should be stated clearly to avoid overclaiming.","section":"Section IV, Fig. 6"}],"minor_comments":[{"comment":"The equation numbering is inconsistent: the capacity expression is labelled 'Equation 2' in the text but is actually the first numbered equation, and subsequent references to 'Equation 2' point to the δmin formula; please renumber the equations.","section":"Section II, Equations"},{"comment":"There are several typos: 'standard division' should be 'standard deviation', 'Eucladian' should be 'Euclidean' in Section II, and 'econonomically' in the Conclusion should be 'economically'.","section":"Section IV and Conclusion"},{"comment":"The column headings 'Average Center Frequencies' and 'Average Bandwidths' are misleading because each cell lists several distinct window values rather than an average; the count column format '4 (≥ 50Ghz), 0( < 50Ghz)' should be explained in the caption.","section":"Table I"},{"comment":"Figure 1 captions should define LA and LT (absorption loss and total path loss) at first use, and the small fonts in the figures reduce readability; Fig. 3 would benefit from a reference to the Six-DoF axes being shown in the bottom-right panel.","section":"Fig. 1 and Fig. 3"}],"recommendation":"major_revision","confidential_remarks":"The manuscript appears to be a lightly expanded version of a workshop-style study. The core idea is timely, but the quantitative results rest on a simulator whose outage model and instance parameters are underspecified; the reported optimal beamwidths (e.g., 23° and 25°) cannot be verified as presented. In addition to the requested revisions, I would encourage the editor to ask the authors to release the simulator code or a detailed pseudocode of the outage and coverage calculation, because without it the claimed numerical optima will remain uncheckable. The scope (parameter modeling for a single AP) is within the journal's remit, but the presentation needs to be brought up to archival standards."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take. The paper's qualitative conclusion—that optimal beamwidth depends on whether users are high-, constrained-, or low-mobility and on where the AP sits—is plausible and probably right. That is the useful part. The static bound in Eq. (2) is a rearrangement of Shannon, so the novelty is the mobility taxonomy and the sweep over AP placements, which is a fair engineering contribution, but not a new principle.\n\nThe trouble is the headline numbers. The paper says outages happen when the body's rotation angle exceeds the beamwidth, but never writes down the exact combination rule. From Table II, S1 peak yaw/pitch/roll are 15.5°, 13.8°, 15°. If any single rotation above δ caused an outage, the optimal δ for S1 should sit just above 15.5°, not at 23–25°. The reported values line up much better with the Euclidean norm of the three rotations (≈25.6°). So the simulation must be using some combination rule, and it is not stated. The same pattern holds for S2 and S3. That makes the central quantitative results unreproducible from the text as written, and it is exactly where the paper's main contribution lives.\n\nOther soft spots: the simulation is underspecified—no room dimensions, user count M, time horizon, or explicit outage equation; the mobility amplitudes are borrowed from unrelated running/VR studies and assigned to service types by hand; and there is no sensitivity analysis. None of these are fatal on their own, but they compound the missing outage rule.\n\nWhat the paper does well: it is clearly organized, it shows the beamwidth dilemma concretely, and it gives the reader a useful mental model of why mobility class and AP placement interact. The related-work citations look appropriate; the paper builds directly on Petrov et al., so the lineage is honest.\n\nWho is this for? Someone doing THz link budget design or beam management who wants a rough sense of how much beamwidth headroom mobility costs. The specific values should not be used until the model is disclosed.\n\nRecommendation: it deserves peer review, not desk rejection. An editor should send it to a referee with a request to pin down the outage model, add the missing simulation parameters, and re-run with sensitivity bounds. If the authors can state their rule and show the optima are robust, the paper becomes a useful engineering reference.","headline":"Useful engineering intuition about THz beamwidth and mobility, but the headline optimal-beamwidth numbers rest on an unstated rotation-combination rule and are not reproducible as written.","tokens_in":9729,"tokens_out":2882,"would_cite":false,"duration_ms":29338,"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":"For moving THz users, beamwidth has a sweet spot","keywords":["Terahertz communication","Indoor mobility model","Adaptive frequency window","Antenna beamwidth","Beamwidth dilemma","Access point placement","Small-scale mobility","User coverage"],"falsifier":"Take the three service classes (VR gaming, walking, sitting), measure real head and torso rotation amplitudes with a motion-capture device across a population, and run the same coverage simulation with those measured amplitudes; if the resulting beamwidth at peak coverage differs by more than a few degrees from the paper's 23 and 25 degree values, or if the outage curve loses its interior maximum, the model's rotation assumptions are unrepresentative. A direct wireless test—sweeping beamwidth on an indoor THz link while a person plays a VR game—would settle the same question empirically.","tokens_in":8623,"feed_emoji":"📡","tokens_out":6224,"duration_ms":53759,"temperature":0.7,"pith_summary":"This paper argues that in indoor terahertz (THz) communication, the antenna beamwidth that maximizes the number of satisfied mobile users is not the narrowest possible beam: for each mobility type and access-point location there is an optimal intermediate beamwidth, a sweet spot between the gain of narrow beams and the outage sensitivity of body rotation. The authors build a mobility model that categorizes users into high, constrained, and low mobility, characterized by a six-dimensional instability parameter tracking position and rotation (yaw, pitch, roll). They simulate a single access point serving multiple users and report specific optimal beamwidths, for example 23 degrees for high-mobility users when the access point is on a wall and 25 degrees when it is on a low table. If correct, this gives network designers a practical rule for adaptively tuning beamwidth and access-point placement to the dominant mobility class in a room, rather than defaulting to the narrowest beam. The paper's value is in converting the 'beamwidth dilemma' from a qualitative concern into a parameterized engineering trade-off.","feed_headline":"For moving THz users, beamwidth has a sweet spot","feed_subtitle":"Simulations show optimal beamwidths shift with mobility type and access-point placement.","key_machinery":"The central mechanism is the 'beamwidth dilemma' expressed through a quantitative trade-off. Antenna gain for a conical main lobe is $G = X/\\delta^2$, with $X$ an aperture-dependent constant, so narrower beams give higher gain; but a link is assumed to go into outage when the user's body rotation angle (yaw, pitch, or roll) exceeds the beamwidth $\\delta$. The mobility model condenses human motion into the system instability parameter $\\Theta$, a six-dimensional vector of position and rotation changes, with per-service-type amplitudes (e.g., S1 uses a $15.5^\\circ$ yaw amplitude, S3 only $5^\\circ$). The simulation also uses a frequency-window lookup table and a minimum-beamwidth bound $\\delta_{\\min}$ derived from the Shannon capacity formula, so each scenario has an outage-limited coverage curve with a clear maximum. That maximum is the paper's optimal beamwidth $\\delta_{\\mathrm{opt}}$, and its location shifts with both the $\\Theta$ statistics of the service type and the access-point geometry.","core_discovery":"The central claim is that for mobile users there exist optimal beamwidths that are affected by the mobility type (high, constrained, low) and access-point placement, and that these optima are computable from a small set of mobility parameters. For a static user, the paper derives a theoretical lower bound on beamwidth from the requested data rate, frequency window, distance, and antenna aperture. For mobile users, it introduces the system instability parameter $\\Theta = [\\Delta x, \\Delta y, \\Delta z, \\Delta \\omega_{\\mathrm{yaw}}, \\Delta \\omega_{\\mathrm{pitch}}, \\Delta \\omega_{\\mathrm{roll}}]$, which captures both translation and rotation of the device, and assigns typical values to three service types: S1 (high mobility, e.g., intense VR gaming), S2 (constrained mobility, fast and slow walking), and S3 (low mobility, sitting/standing). Using a random-waypoint simulation with these $\\Theta$ patterns and three access-point placements (ceiling, wall, and table), the paper finds that while throughput generally rises as the access point moves closer to users, user coverage is maximized at an intermediate beamwidth: for S1 users the optimum is $\\delta_{\\mathrm{opt}} = 23^\\circ$ for wall placement and $\\delta_{\\mathrm{opt}} = 25^\\circ$ for table placement, while S2 and S3 reach their coverage optima at lower beamwidths. This directly addresses the beamwidth dilemma: narrower beams give higher gain but suffer outages when body rotation exceeds the beamwidth, so the best beamwidth balances gain against rotation-induced misalignment.","pith_inferences":["If the mobility amplitudes were measured directly from real users in each service class, the $\\delta_{\\mathrm{opt}}$ values would likely shift, but the qualitative structure—an interior optimum set by the balance of gain and rotation-induced outage—should persist for any unimodal rotation distribution.","The same $\\Theta$-based outage model could be applied to reconfigurable intelligent surfaces or multi-AP cooperative systems, where the optimal beam pattern might be a function of the dominant instability direction rather than a single scalar $\\delta$.","A natural field test would be to run a VR session with an indoor THz link and sweep beamwidth while logging packet errors; the resulting coverage curve should peak near the predicted $\\delta_{\\mathrm{opt}}$ if the model's rotation amplitudes are representative.","The paper's static minimum-beamwidth bound suggests a quick heuristic: choose the smallest beamwidth that satisfies the rate requirement at the maximum expected user distance, then broaden it just enough to cover the largest expected rotation amplitude."],"forward_implications":["An indoor THz access point can improve the number of satisfied users by dynamically steering beamwidth to the $\\delta_{\\mathrm{opt}}$ of its dominant mobility class instead of always using the narrowest beam.","Wall and table placements are better than ceiling placement for capturing yaw and pitch movements, so single-AP deployments for walking or sedentary users should favor lower, closer antenna positions.","For high-mobility services like VR gaming, coverage is maximized at beamwidths around 23 to 25 degrees, not at single-digit beamwidths which would suffer frequent rotation outages.","Frequency-window selection and beamwidth tuning are interdependent: the static minimum beamwidth bound from target data rate and distance sets a floor, and mobility pushes the operating point above that floor.","The service-type classification (S1/S2/S3) with its $\\Theta$ parameter set can serve as a control input for adaptive beamforming in future THz systems."],"supporting_citations":[{"why":"Establishes that small-scale mobility causes outages in THz links and motivates the beamwidth dilemma.","marker":"[11]"},{"why":"Provides the aperture constant X used in the antenna gain formula $G = X/\\delta^2$.","marker":"[14]"},{"why":"Supplies the bobbing amplitude for walking (Δz) used in the constrained-mobility service type.","marker":"[15]"},{"why":"Provides the sinusoidal yaw, pitch, and roll patterns and amplitudes for walking.","marker":"[16]"},{"why":"Source for the high-mobility (S1) rotation amplitudes based on human running.","marker":"[17]"},{"why":"Gives position and orientation noise values for the VR headset used in S1.","marker":"[18]"},{"why":"Provides the atmospheric attenuation model that shapes the frequency-window lookup table.","marker":"[4]"},{"why":"Supplies the capacity formula used to compute achievable data rate and minimum beamwidth.","marker":"[13]"}],"fun_headline_variants":["Optimal THz beamwidth scales with mobility and access point","Beamwidth sweet spot in THz shifts with user mobility type","Narrow beams aren't always best: THz coverage has a peak","Indoor THz: choose beamwidth by mobility, not just gain","Mobility-aware beamwidth selection boosts THz coverage"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a link outage occurs whenever a user's body rotation (yaw, pitch, or roll) exceeds the antenna beamwidth, and that the rotation amplitudes assigned to each service type—borrowed from separate motion studies of running, walking, and VR tracking—accurately represent indoor user behavior.","fun_headline_variants_meta":{"raw":{"variants":["Optimal THz beamwidth scales with mobility and access point","Beamwidth sweet spot in THz shifts with user mobility type","Narrow beams aren't always best: THz coverage has a peak","Indoor THz: choose beamwidth by mobility, not just gain","Mobility-aware beamwidth selection boosts THz coverage"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000496,"raw_usage":{"total_tokens":2492,"prompt_tokens":1062,"completion_tokens":1430,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":678,"completion_tokens_details":{"reasoning_tokens":1341}},"tokens_in":678,"tokens_out":1430,"duration_ms":10892,"temperature":1.0,"reasoning_tokens":1341,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:23:28.761240+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the three service classes (VR gaming, walking, sitting), measure real head and torso rotation amplitudes with a motion-capture device across a population, and run the same coverage simulation with those measured amplitudes; if the resulting beamwidth at peak coverage differs by more than a few degrees from the paper's 23 and 25 degree values, or if the outage curve loses its interior maximum, the model's rotation assumptions are unrepresentative. A direct wireless test—sweeping beamwidth on an indoor THz link while a person plays a VR game—would settle the same question empirically.","supporting_citations":[{"cited_title":"The effect of small-scale mobility on terahertz band communications,","cited_arxiv_id":null,"evidence_quote":"Establishes that small-scale mobility causes outages in THz links and motivates the beamwidth dilemma."},{"cited_title":"Gain of Directional Antennas,","cited_arxiv_id":null,"evidence_quote":"Provides the aperture constant X used in the antenna gain formula $G = X/\\delta^2$."},{"cited_title":"The up and do wn bobbing of human walking: a compromise between muscle work and effici ency","cited_arxiv_id":null,"evidence_quote":"Supplies the bobbing amplitude for walking (Δz) used in the constrained-mobility service type."},{"cited_title":"Interaction o f the body, head, and eyes during walking and turning,","cited_arxiv_id":null,"evidence_quote":"Provides the sinusoidal yaw, pitch, and roll patterns and amplitudes for walking."},{"cited_title":"Upper-Body Control and Mechanism of H umanoids to Compensate for Angular Momentum in the Y aw Direction Based o n Human Running,","cited_arxiv_id":null,"evidence_quote":"Source for the high-mobility (S1) rotation amplitudes based on human running."},{"cited_title":"The Accuracy and Prec ision of Position and Orientation Tracking in the HTC Vive Virtual Re ality System for Scientific Research,","cited_arxiv_id":null,"evidence_quote":"Gives position and orientation noise values for the VR headset used in S1."},{"cited_title":"Attenuation by atmospheric gases-P Series Radiowa ve propaga- tion,","cited_arxiv_id":null,"evidence_quote":"Provides the atmospheric attenuation model that shapes the frequency-window lookup table."},{"cited_title":"Improving User Cov erage Through Resource Aware Handoff Management in Heterogeneou s Net- works,","cited_arxiv_id":null,"evidence_quote":"Supplies the capacity formula used to compute achievable data rate and minimum beamwidth."}],"review_version":1}