{"id":"075be575-f6f2-4e2c-9e27-7dddb98d2e5f","arxiv_id":"2607.29288","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"In a Wi-Fi 8-inspired MU-MIMO model, adding a Glaze-style batteryless overlay can shift the throughput-optimal channel-sounding interval, usually to shorter values at high MCS.","lead":"This paper models a Wi-Fi 8-like system in which a battery-free receiver piggybacks data onto the access point's beamformed MU-MIMO signal, and asks how often the AP should re-sound the channel. In simulations, the passive overlay shifts the optimal sounding interval, usually to shorter values at high modulation rates, but the result depends on unverified model choices.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Unvalidated link-level functions and unreleased code make the claimed 53 ms vs 59 ms optimum shift irreproducible; the result may be an artifact of f_W, f_P, and xi_W.","rationale":"The reader's weakest assumption correctly identifies the unspecified nature of f_W, f_P, and xi_W as the core weakness. My stress test confirms that this is the single most load-bearing issue: the paper's only quantitative evidence for the central claim is a small shift in the optimal sounding interval, and that shift is entirely dependent on unvalidated, undisclosed modeling functions. Without these functions or the simulator, the result cannot be independently reproduced or assessed, so it is impossible to determine whether the effect is physical or an artifact. The paper's theoretical framework (Section II-F) is plausible, and the data-driven search procedure is reasonable, but the numerical evidence is not sufficient to support a stronger verdict than CONDITIONAL. I do not recommend changing the reader's verdict because the issue is addressable (e.g., by releasing code and specifying the functions) rather than a demonstrated internal inconsistency. The concern about statistical significance (lack of error bars) is secondary but reinforces the need for caution. Therefore, the verdict remains CONDITIONAL, and UNCHANGED is appropriate.","tokens_in":8888,"tokens_out":6911,"duration_ms":75156,"concrete_test":"Provide the exact functional forms and parameter values for f_W, f_P, and xi_W used in the simulations and release the simulator. Then recompute T_p* for MCS 4 across a grid of xi_W (e.g., 0.1 to 10) and Delta (e.g., 1-6 dB) with 100 independent Monte Carlo runs, reporting 95% confidence intervals. If the passive-enabled and baseline T_p* intervals overlap, or if the shift reverses sign for a plausible parameter value, the central claim is not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing assertion is the MCS-4 result that the passive overlay shifts the optimal post-sounding interval from 59 ms to 53 ms. This number depends on the packet-error functions f_W (Eq. 15) and f_P (Eq. 19), and the switching-distortion coefficient xi_W (Eq. 14). None of these are specified, derived from first principles, or calibrated against measurements. The paper does not release the MATLAB simulator or the exact parameter values (Delta, R_b, Omega_P, sigma_P^2) used in the figures. Consequently, the 6 ms shift, which is only about 10% of the optimum, cannot be distinguished from an artifact of arbitrary modeling choices. The absence of error bars or confidence intervals on T_p* compounds this: the training-validation gap of 1 ms is reported for the chosen operating condition, but no statistical test shows that the passive and baseline optima are significantly different. If a different but equally reasonable choice of f_W or xi_W were used, the shift could disappear or reverse, undermining the central claim that passive overlay 'fundamentally changes' the sounding tradeoff.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a Wi-Fi 8/802.11bn-inspired downlink MU-MIMO system with a Glaze-style batteryless overlay. It formulates a cycle-average throughput objective that includes sounding overhead, CSI aging, Wi-Fi and passive packet errors, and passive embedding feasibility, and then optimizes the post-sounding interval T_p for fixed MCS and passive-overlay parameters. The main numerical claim is that at MCS 4 the passive overlay shortens the optimal post-sounding interval from 59 ms to 53 ms, and that the MCS sweep separates into an overhead-limited low-MCS regime and an aging-limited high-MCS regime. The paper also gives a first-difference characterization of the optimum and a unimodality argument under monotone aging.","tokens_in":9260,"tokens_out":6616,"duration_ms":67315,"significance":"If the numerical result could be trusted, the paper would be a useful extension of prior sounding-interval optimization to passive-enabled MU-MIMO WLANs. The analytical component that is actually shown is mostly sound: the cycle-average formulation in Eq. (25), the one-step condition in Eq. (30), and the unimodality argument under monotone packet-throughput aging are correct. The data-driven search in Algorithm 1 is simple and has linear complexity. However, the load-bearing packet-error functions, the switching-distortion model, the passive scalar-equivalent model, and several key parameters are left unspecified or unvalidated, so the central quantitative claim is currently not reproducible. The paper would be significantly strengthened by specifying the full link-level model, releasing the simulator, and adding a statistical comparison of the baseline and passive optima.","major_comments":[{"comment":"The central result—the passive overlay shortens T_p* from 59 ms to 53 ms at MCS 4—is produced by a MATLAB link simulator whose packet-error functions f_W and f_P, the switching-distortion coefficient ξ_W, and the passive-link constants (Δ, R_b, Ω_P, σ_P^2, K_c, T_P^pre) are never specified. The text only says an HE/TGax link-level model is used. No code or data is released. The 6 ms shift is about 10% of the optimum, and no confidence intervals or significance test is provided; the single reported 1 ms training-validation gap at one operating point does not establish that 53 ms and 59 ms are statistically distinguishable. The claimed regime change could be an artifact of the unspecified modeling choices. Please provide the functional forms, all parameter values, a sensitivity analysis around the chosen parameters, and a statistical comparison of the baseline and passive optima.","section":"§III-A, Figs. 2–3; Eqs. (14), (15), (19)"},{"comment":"The Wi-Fi loss induced by amplitude switching is modeled as D_ovl = ξ_W(R_b,m_k) σ_a^2 (S+I), with no derivation. This is not a standard result for amplitude-modulated overlay with coherent reception; it is an added distortion power whose coefficient ξ_W is not defined, tabulated, or calibrated. Because this term is one of the two mechanisms that can make the passive system prefer shorter sounding intervals, the direction and magnitude of the predicted shift are directly controlled by an unvalidated assumption. Please derive D_ovl from a signal model or calibrate ξ_W with link-level measurements; otherwise the conclusion in §III-A cannot be separated from this assumption.","section":"§II-C, Eq. (14)"},{"comment":"The passive link is modeled by a scalar-equivalent channel and envelope detector with γ_P = F_s(1−a_Δ)^2 Ω_P[n]/(R_b σ_P^2) and packet-error function f_P. No justification is given for why the full MU-MIMO waveform can be reduced to a scalar x[n,t], nor are F_s, Ω_P[n], σ_P^2, the passive payload length, or the threshold model defined. The passive throughput contribution D_P in Eq. (23) is load-bearing for the claimed shift, so this lack of specification is not merely cosmetic. Provide the derivation of the scalar-equivalent model and all numerical constants, or replace the passive model with a measured/validated packet-error curve.","section":"§II-C, Eqs. (16)–(19)"},{"comment":"Eq. (35) gives a necessary condition for a left shift, but the paper does not verify it in simulation. The verbal explanation that the Wi-Fi loss δ_W is strongest at aged CSI is plausible, but Fig. 2 reports only aggregate curves; showing the empirical Δz[n] and the baseline/passive cycle averages near p_B* would directly test Eq. (35). Without such a decomposition, the link between the theoretical condition and the 53 vs 59 ms result remains assertive. Please add a plot or table of E[z_tot[p_B*]] versus Rsum(p_B*−1) for both systems.","section":"§II-F, Eq. (35) and §III-A"}],"minor_comments":[{"comment":"Symbols F_s, Ω_P[n], σ_P^2, K_c, T_P^pre, and the passive packet length are not defined. Please define all symbols and give the numerical values used in the simulations.","section":"§II-C, Eq. (18)-(21)"},{"comment":"The sentence citing [10]–[12] for 'Wi-Fi backscatter and Passive Wi-Fi' includes two references—a 6G C-V2X radio environment map paper and an AI programming-assistants paper—that appear unrelated to backscatter. Please correct the citation list.","section":"References [11], [12]"},{"comment":"The table lists MCS range, sounding overhead, and channel model, but omits the passive overlay parameters Δ and R_b, and the antenna/user configuration is only in the text. All simulation parameters needed to reproduce Figs. 2–4 should be in the table or in a dedicated appendix.","section":"Table I"},{"comment":"Fig. 4 is described only qualitatively ('rapidly decreases'). Please report the numerical values for the MCS sweep, and add error bars or shaded confidence bands to Figs. 2 and 3.","section":"Figures 2–4"},{"comment":"The tie-breaking rule when two values of p give the same maximum is not specified. Since T_p* is reported to the millisecond, please state how ties are resolved.","section":"Algorithm 1"}],"recommendation":"major_revision","confidential_remarks":"The paper's main quantitative claim is contingent on undisclosed simulator internals and unspecified link-level functions, so the current version is not reproducible. In addition, the reference list contains two apparently unrelated entries ([11], [12]) in the backscatter citation cluster, which I would ask the editor to check for citation integrity. If the authors can provide the full model details, parameter tables, and the simulator, I would be willing to reconsider."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [Colleague],\n\nThe paper is a coherent extension of [8] to a Glaze-style passive overlay. The new piece is the coupling: the overlay's attenuation and switching distortion interact with CSI aging, so the optimal sounding interval can shift. The theory in Section II-F is clean: the first-difference condition and the unimodality argument under monotone aging are correct, and the overhead-limited/aging-limited regime split is useful. The simulations at MCS 4 show a 53 ms optimum for the passive-enabled system versus 59 ms for the baseline—a plausible, self-sufficient result if you trust the model.\n\nBut that \"if\" is where I get stuck. The load-bearing packet-error functions f_W (Eq. 15) and f_P (Eq. 19), and the switching-distortion coefficient xi_W (Eq. 14), are unspecified. The simulator and exact parameter values (Delta, R_b, Omega_P, sigma_P^2) are not released. So the 6 ms shift is not independently checkable. Worse, there are no error bars or hypothesis tests showing the two optima are significantly different. The training-validation gap of 1 ms does not address the difference between baseline and passive cases.\n\nThere is also a mild circularity: the model builds in D_ovl that grows with aged CSI, so the simulation \"discovers\" that the passive overlay shortens the interval. The right-shift case mentioned in theory is never shown. So the claim that \"the passive overlay fundamentally changes the sounding tradeoff\" is overstated for what is a 10% left-shift under one MCS.\n\nStill, I don't read this as a dishonest paper. It is a clean modeling exercise with explicit assumptions. The issue is that the numerical claims are presented as findings rather than as illustrations of the model. For a design-guidance paper, you need either real link-level calibration or a release of the simulator so others can probe the sensitivity. With that, it could be a useful reference for Wi-Fi 8 sounding design.\n\nI'd send it to a serious referee, but with a firm request to specify or calibrate the omitted functions, show the right-shift scenario, and provide error bars on the optimum.\n\nBest.","headline":"Clean model and theory, but the headline 53 vs 59 ms shift rests on uncalibrated link functions and no error bars—needs a rework before the numbers are trusted.","tokens_in":9662,"tokens_out":2733,"would_cite":true,"duration_ms":29877,"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":"This paper shows that adding a batteryless passive link to a Wi-Fi 8 downlink MU-MIMO system changes the optimal channel-sounding interval, so fixed-period sounding is no longer sufficient.","keywords":["channel sounding","MU-MIMO","Wi-Fi 8","IEEE 802.11bn","batteryless communication","ambient backscatter","Glaze overlay","CSI aging"],"falsifier":"Measure the optimal post-sounding interval on an SDR testbed with a real Glaze-style envelope detector and a 4-antenna AP, sweeping T_p at MCS 4; if the throughput maximum lands within a few ms of the no-overlay baseline and does not move left, the claimed reshaping of the tradeoff is not supported.","tokens_in":8794,"feed_emoji":"📡","tokens_out":3753,"duration_ms":38347,"temperature":0.7,"pith_summary":"The paper tries to establish that a batteryless overlay that embeds passive bits into the Wi-Fi downlink amplitude does not merely add a constant throughput; it couples to the channel sounding cycle and reshapes the tradeoff between sounding overhead and CSI aging. A packet-level cross-layer model computes cycle-average throughput, and a data-driven search finds the post-sounding interval that maximizes it under each MCS and passive configuration. Simulations show that at MCS 4 the optimal interval shifts from 59 ms to 53 ms when the passive overlay is present, and that lower MCS levels remain overhead-limited while higher MCS levels become aging-limited. The consequence is that fixed-period sounding is suboptimal for passive-enabled MU-MIMO WLANs, and the interval should be adapted jointly with MCS and passive settings.","feed_headline":"Passive IoT overlay shifts optimal Wi-Fi sounding interval","feed_subtitle":"At MCS 4, the best post-sounding interval drops from 59 ms to 53 ms, so fixed-period sounding falls short.","key_machinery":"The key machinery is the cycle-average aggregate throughput Rsum(p; theta) = (sum of expected delivered Wi-Fi and passive payload bits over p packets) / (T_d + p T_0), together with the one-step difference condition that determines when to stop extending the interval. The Glaze-style amplitude overlay is the central physical ingredient: passive bits are Manchester-encoded into two amplitude states controlled by attenuation depth Delta, and the same scalar factor is applied across all spatial streams so beamforming direction is preserved. The data-driven search evaluates all candidate p values in O(P_max) operations by recursive accumulation of the empirical per-packet throughput averaged ove","core_discovery":"The central discovery is that the passive overlay's attenuation depth and embedding rate interact with CSI aging: the same amplitude distortion that carries passive bits becomes more costly as the precoder ages, so the passive stream can shift the throughput-optimal sounding interval left or right depending on MCS and passive-link parameters. The paper derives a one-step condition: extending the post-sounding stage is beneficial exactly while the next packet's expected throughput exceeds the current cycle average. In the MCS-4 simulation, the passive-enabled optimum drops from 59 ms to 53 ms, because late-cycle packets suffer from both outdated precoding and overlay distortion, while the pas","pith_inferences":["If the unvalidated packet-error and switching-distortion models are replaced by empirical lookup tables from a real 802.11bn-like receiver, the same framework could be recalibrated without changing the optimization machinery.","The left-shift at MCS 4 suggests an online policy could trigger an extra sounding whenever the observed per-packet throughput of a late packet falls below the running cycle average, rather than on a fixed timer.","In deployments where passive traffic dominates, tuning the attenuation depth to keep the passive stream robust as CSI ages could preserve long sounding intervals and lower sounding overhead.","The same one-step condition applies to any periodic overhead-vs-freshness tradeoff, e.g., beam training in mmWave or tracking in massive MIMO, whenever the per-slot goodput is non-increasing with age."],"forward_implications":["Fixed-period sounding is suboptimal once a passive overlay is present; the optimal post-sounding interval must be re-derived for each MCS and passive configuration.","At low MCS (0–3) the system is overhead-limited and the optimum sits at the 180 ms upper bound, so long sounding intervals are safe; at MCS 4–9 aging dominates and the optimum shrinks.","The one-step decision rule holds: keep extending the data stage exactly until the next packet's expected throughput drops below the current cycle average.","In the MCS-4 simulation, adding the passive overlay moves the optimum from 59 ms to 53 ms, showing the overlay can justify more frequent CSI refresh rather than a constant throughput bonus.","Estimating the throughput profile from K=30 training realizations yields an optimum within 1 ms of the validation optimum, indicating the profile is smooth enough for practical interval selection."],"fun_headline_variants":["Passive overlay shifts Wi-Fi sounding sweet spot","Batteryless links alter optimal Wi-Fi sounding interval","Wi-Fi 8 sounding: passive overlay changes the tradeoff","At MCS 4, passive overlay cuts best interval to 53ms","Passive overlay reshapes Wi-Fi sounding tradeoff"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The quantitative shift of the optimal interval rests on unvalidated packet-error functions and a single simulated 0.089 m/s indoor channel; if real Wi-Fi 8 and batteryless receivers behave differently, the 53-versus-59 ms shift may not appear.","fun_headline_variants_meta":{"raw":{"variants":["Passive overlay shifts Wi-Fi sounding sweet spot","Batteryless links alter optimal Wi-Fi sounding interval","Wi-Fi 8 sounding: passive overlay changes the tradeoff","At MCS 4, passive overlay cuts best interval to 53ms","Passive overlay reshapes Wi-Fi sounding tradeoff"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000513,"raw_usage":{"total_tokens":2311,"prompt_tokens":710,"completion_tokens":1601,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":454,"completion_tokens_details":{"reasoning_tokens":1518}},"tokens_in":454,"tokens_out":1601,"duration_ms":12917,"temperature":1.0,"reasoning_tokens":1518,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T09:57:50.325485+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the optimal post-sounding interval on an SDR testbed with a real Glaze-style envelope detector and a 4-antenna AP, sweeping T_p at MCS 4; if the throughput maximum lands within a few ms of the no-overlay baseline and does not move left, the claimed reshaping of the tradeoff is not supported.","supporting_citations":[],"review_version":1}