{"id":"6df7e3d3-60e1-4374-8fe6-e5f6ea0dd1c4","arxiv_id":"2607.28213","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":3,"one_line_summary":"Conducted tests find no 5G NR degradation from VLP Wi-Fi 6E/7 below −75 dBm, and the implied harm radius sits well inside the LBT exclusion zone.","lead":"Lab tests show a commercial Wi-Fi 6E/7 gadget does not hurt a live 5G link at 6 GHz until its signal is quite strong (−75 dBm). Mapping that level to distance puts the danger zone well inside the area where Wi-Fi must already stay silent, so rules already on the books appear to protect 5G.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified beyond the reader's already-flagged LBT/propagation conditionality.","rationale":"The reader's strongest_claim accurately separates the lab result (−75 dBm onset on both receivers) from the link-budget wrap. The weakest_assumption correctly isolates the imported ECC/ITU parameters and real-device ED behavior as the point on which the 4.9–6.1× margin can shrink. That is already the right load-bearing concern; the experimental campaign itself is sound for a conducted first-of-kind dual-receiver measurement, the limitations section is candid, and no internal inconsistency, arithmetic error, or unstated assumption inside the measurement chain rises to the same level. Consequently the CONDITIONAL verdict (accept the threshold; treat the spatial-protection slogan as conditional on the stated propagation/ED premises) needs no adjustment. The concrete_test simply stress-tests the already-flagged parameters rather than introducing a new objection.","tokens_in":10478,"tokens_out":552,"duration_ms":10474,"concrete_test":"Recompute d_LBT (Eqs. 3–4) under a 10 dB reduction in effective broadcast EIRP (e.g., 56 dBm/20 MHz) or L_EX=15 dB while holding the measured −75 dBm onset fixed; if the resulting exclusion zone falls below ~112 m the 4.9–6.1× margin disappears and the coexistence slogan requires qualification, otherwise the reader's CONDITIONAL stance is reinforced.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central experimental claim (no measurable degradation of either NR receiver below −75 dBm injected commercial VLP power) is directly supported by the conducted sweeps in §V and Figs. 3–4. The coexistence slogan that a compliant VLP therefore cannot harm either receiver rests on the analytical mapping in §III (Eqs. 2–5), which the reader already correctly identifies as the weakest link: it imports ECC/ITU parameters (66 dBm/20 MHz broadcast, L_EX≈23 dB urban NLoS, free-space harm path, −62 dBm ED) and assumes real devices defer on SSB energy. Those assumptions are external and standard for the policy setting the paper addresses; they are not internal contradictions, hidden algebraic errors, or unacknowledged gaps. The authors themselves list single-device continuous injection, static LOS, and missing aggregate/adjacent-channel cases as limitations (§VI-D) and future work. No stronger load-bearing flaw is present that would overturn the measured threshold or the conditional character of the spatial-margin claim.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper reports the first conducted co-channel interference measurements of a commercial VLP Wi-Fi 6E/7 access point into both the gNB uplink and UE downlink receiver chains of a live 5G NR system in band n102. Using an OpenAirInterface/O-RAN SDR testbed with full 5G core, the authors sweep injected Wi-Fi power and record throughput, BLER, and SNR. Neither receiver shows measurable degradation below −75 dBm; above that threshold performance degrades progressively, with the UE more resilient at lower MCS and under beacon-only traffic. A log-distance link-budget analysis then maps the measured onset to a free-space harm radius of ~112 m, which lies inside the 545–685 m LBT exclusion zone derived from ECC/ITU parameters, supporting the claim that a compliant VLP device would defer before harming either receiver.","tokens_in":10699,"tokens_out":1207,"duration_ms":26503,"significance":"The work fills a clear gap: prior 6 GHz coexistence studies are largely analytical or simulation-based and do not characterise real Wi-Fi 6E/7 hardware against both gNB UL and UE DL chains of an end-to-end NR stack. The dual-victim power sweeps, commercial interferer, and consistent metrics (throughput/BLER/SNR) provide a concrete empirical anchor (−75 dBm onset) that spectrum-policy discussions can use. The explicit mapping of lab thresholds to LBT versus harm radii is a useful bridge to ECC Report 366-style sharing studies. Strengths include the complete OAI core-plus-RAN setup, baseline B0 controls, and transparent limitations section. If the conditional spatial-margin claim holds under the stated propagation and ED assumptions, the result is directly relevant to regulators and operators evaluating VLP–MFCN coexistence in n102/n104.","major_comments":[{"comment":"The abstract and §VI–VII present the spatial-protection conclusion (“a compliant VLP device would vacate the channel before its emissions could harm either receiver”) as confirmed by the measurements. That conclusion rests on the analytical model in §III (Eqs. 2–5), which imports free parameters—macro broadcast EIRP of 66 dBm/20 MHz, L_EX ≈ 23 dB from ITU-R P.1411-12 urban NLoS for the detection path, free-space (L_EX = 0) for the harm path, and reliable deferral at −62 dBm ED on SSB/broadcast energy. The measured −75 dBm onset is solid; the 4.9–6.1× margin is not. The manuscript should state the coexistence claim as conditional on these standard but external assumptions, and should quantify sensitivity (e.g., how the margin changes if L_EX drops by 5–10 dB or if effective broadcast EIRP is lower).","section":"§III, Eqs. (2)–(5); Abstract; §VI-C"},{"comment":"§IV-C and §V inject continuous heavy Wi-Fi traffic while bypassing the VLP’s own LBT. This is a legitimate worst-case method for finding the raw receiver damage threshold and is acknowledged in §VI-D, yet the paper never reports the actual duty cycle or channel occupancy achieved during the sweeps. Without that number, it is hard to judge how far the continuous-injection case sits from the 3–5 % duty-cycle figure used later in the discussion. A short occupancy measurement (spectrum-analyser or airtime statistic) at the operating points near −75 dBm would strengthen the link between lab results and the claimed real-world exposure reduction.","section":"§IV-C; §V; §VI-D"}],"minor_comments":[{"comment":"Fig. 4 captions incorrectly label panels (a) and (c) as “5G NR BLER”; (a) is SNR and (c) is throughput. Correct the legend text.","section":"Fig. 4"},{"comment":"Abstract and several body passages write “75 dBm” without the minus sign; the intended threshold is −75 dBm. Fix consistently.","section":"Abstract"},{"comment":"Table I lists UE pMax as 20 dBm while §VI-C later invokes 23 dBm EIRP (minus body loss) for secondary LBT protection. Align the UE power figures or note the regulatory versus testbed distinction.","section":"Table I; §VI-C"},{"comment":"The TDD pattern (7 DL + 2 UL slots) makes the UL measurement interval short; a brief note on how many UL TTIs were averaged for the SNR/BLER points in Fig. 4 would aid reproducibility.","section":"§V-D; Fig. 4"},{"comment":"Reference [9] (ECC Report 366) is dated Jun. 2025 in the bibliography; confirm the public citation details once the report is final.","section":"References"}],"recommendation":"minor_revision","confidential_remarks":"The experimental core is sound and novel enough for a solid contribution in the coexistence-measurement literature. The main risk is over-claiming in the abstract relative to the conditional analytics; once the authors soften that language and add a short sensitivity note, the paper should be acceptable. Fit for a networking/spectrum journal is good; less so for a pure theory venue."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The thing worth knowing is the lab result: they actually injected a commercial TP-Link VLP Wi-Fi 6E/7 AP into both the gNB UL and UE DL chains of a live OAI/O-RAN stack with core in n102, swept power, and saw no measurable TP/BLER/SNR hit below −75 dBm on either receiver. Progressive degradation above that, UE more resilient at 60 Mbit/s than 85, and beacon-only essentially harmless up to −55 dBm. That hardware gap relative to the stochastic-geometry, ns-3, and adjacent-channel papers they cite is real.\n\nThey do the experiment cleanly. Baseline B0, two DL rates plus beacon-only, gNB UL with SNR/BLER/TP, conducted architecture with spectrum-analyser checks, configs detailed enough to follow. Limitations section is honest: LBT bypassed, single static source, no aggregate/adjacent/mobility. The link-budget wrap (harm radius ~112 m free-space vs 545–685 m LBT exclusion) is ordinary path-loss algebra using ECC 366 EIRP, ITU-R P.1411 L_EX ≈ 23 dB, and −62 dBm ED; it does not invent the threshold from the desired margin. Circularity is low.\n\nSoft spots are proportionate, not load-bearing. No error bars or repeated trials, no public traces/code, and the coexistence slogan (“compliant VLP vacates before it can harm”) inherits whatever real-world EIRP, excess loss, or device ED behaviour actually is. If those differ, the 4.9–6.1× margin shrinks while the −75 dBm lab number stays intact. Authors already flag the single-device continuous-injection worst case. Math and citations look fine; nothing incoherent.\n\nThis is for people working 6 GHz coexistence, spectrum policy, or NR-U/Wi-Fi sharing who need a concrete measured onset rather than another simulation. I would cite the threshold and the dual-receiver setup. It deserves a serious referee—accept the measurement, keep the spatial-margin claim conditional. Worth a look; not a must-read outside the subfield.","headline":"Solid first conducted dual-receiver measurement of commercial VLP Wi-Fi into live 5G NR in n102; the −75 dBm onset is the real result, and the LBT spatial-margin claim is standard link-budget wrapping that stays conditional on ECC/ITU parameters.","tokens_in":11441,"tokens_out":581,"would_cite":true,"duration_ms":16832,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"A compliant VLP Wi-Fi 6E/7 device vacates the 6 GHz channel before its power can degrade either 5G NR receiver.","keywords":["5G NR","Wi-Fi 6E","Wi-Fi 7","VLP","coexistence","LBT","band n102","O-RAN"],"falsifier":"Place a commercial VLP access point at a known distance between 112 m and 545 m from an operating macro gNB, verify that the VLP still senses the SSB/broadcast above −62 dBm and defers, and confirm that neither gNB uplink nor UE downlink metrics degrade when the VLP is forced to transmit at full power.","tokens_in":11298,"feed_emoji":"📡","tokens_out":1013,"duration_ms":17735,"temperature":0.7,"pith_summary":"This paper reports the first conducted measurements of a commercial very-low-power Wi-Fi 6E/7 device interfering into both the uplink of a live 5G base station and the downlink of a user device, using a full software-defined 5G stack in the 6 GHz band. Neither receiver shows measurable loss of throughput, rise in block errors, or drop in signal quality until the injected Wi-Fi power exceeds −75 dBm. Mapping that laboratory threshold through a simple link budget yields a harm radius of roughly 112 m, which sits well inside the 545–685 m zone in which the base station’s own broadcasts force the Wi-Fi device to defer. The practical claim is therefore that ordinary Listen-Before-Talk already protects both 5G receivers; no extra mitigation is required for a single compliant VLP device. A sympathetic reader cares because regulators and operators need hardware evidence, not only simulations, before they can trust shared use of the newly opened 6 GHz band.","feed_headline":"Wi-Fi at 6 GHz stays outside the 5G harm zone","feed_subtitle":"Lab threshold of −75 dBm maps to 112 m; Listen-Before-Talk already clears 545 m","key_machinery":"The spatial-margin argument that compares the measured −75 dBm degradation onset (converted to a free-space harm radius of ~112 m) against the analytically derived LBT exclusion radii of 545–685 m obtained by inverting a log-distance path-loss model at the −62 dBm energy-detection threshold.","core_discovery":"Neither the gNB uplink nor the UE downlink receiver of a live 5G NR system in band n102 exhibits measurable degradation below −75 dBm of injected commercial VLP Wi-Fi power. The free-space distance at which a 14 dBm VLP reaches that threshold is about 112 m, while the same VLP must defer once it is inside the 545–685 m Listen-Before-Talk exclusion zone created by the base-station broadcast. Consequently a standards-compliant VLP device leaves the channel before its emissions can harm either receiver.","pith_inferences":["If real urban excess loss on the detection path is lower than the 23 dB median used here, the exclusion zone shrinks and the safety margin could fall below the reported 5× factor.","Aggregate interference from many co-located VLP devices could push the effective harm radius outward even while each individual device still obeys LBT, a regime the single-source testbed does not address.","The same conducted-threshold approach could be used to set evidence-based ED thresholds for future 6 GHz or 7 GHz sharing rules rather than relying solely on simulation."],"forward_implications":["A single compliant VLP device needs no additional coexistence mitigation beyond existing LBT in band n102.","The −62 dBm energy-detection threshold already provides a several-fold spatial safety margin under the paper’s link-budget assumptions.","Lower-MCS 5G links and beacon-only Wi-Fi traffic are substantially more resilient, so light-load or control-plane-only scenarios are even safer.","The same measurement-plus-link-budget method can be repeated for adjacent-channel and multi-device cases to test whether the margin survives aggregation."],"fun_headline_variants":["VLP Wi-Fi spares 5G NR below -75 dBm at 6 GHz","5G NR uplink and downlink hold firm under -75 dBm Wi-Fi","-75 dBm Wi-Fi threshold sits inside 545 m LBT exclusion","Compliant VLP vacates 6 GHz before harming 5G receivers","UE and gNB show no loss until Wi-Fi exceeds -75 dBm"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The claim that the Wi-Fi device always stays outside the harm zone rests on the calculated size of the Listen-Before-Talk exclusion radius, which itself depends on assumed base-station broadcast power, urban excess loss, and that real devices actually defer at the regulatory energy-detection threshold.","fun_headline_variants_meta":{"raw":{"variants":["VLP Wi-Fi spares 5G NR below -75 dBm at 6 GHz","5G NR uplink and downlink hold firm under -75 dBm Wi-Fi","-75 dBm Wi-Fi threshold sits inside 545 m LBT exclusion","Compliant VLP vacates 6 GHz before harming 5G receivers","UE and gNB show no loss until Wi-Fi exceeds -75 dBm"]},"model":"grok-4.5","effort":"low","cost_usd":0.005186,"raw_usage":{"total_tokens":1450,"prompt_tokens":824,"num_sources_used":0,"completion_tokens":94,"cost_in_usd_ticks":51864000,"prompt_tokens_details":{"text_tokens":824,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":532,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":824,"tokens_out":94,"duration_ms":8770,"temperature":1.0,"reasoning_tokens":532,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-31T14:27:00.743676+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Place a commercial VLP access point at a known distance between 112 m and 545 m from an operating macro gNB, verify that the VLP still senses the SSB/broadcast above −62 dBm and defers, and confirm that neither gNB uplink nor UE downlink metrics degrade when the VLP is forced to transmit at full power.","supporting_citations":[],"review_version":1}