{"id":"fe8e2cd6-6095-48b2-aa40-11d9812bace1","arxiv_id":"1908.05946","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Vehicle-mounted millimeter-wave relays can more than double average user spectral efficiency in dense pedestrian, moderate traffic urban streets when users can instantly pick the best link.","lead":"This paper builds a mathematical model of a city street where some cars act as moving signal relays for millimeter-wave 5G networks, and shows these relays can more than double a user's connection speed in crowded areas. It helps network designers predict where vehicle-mounted relays will help before deploying them.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Aggressive-relay SE in Eq. (21) is set equal to the COW-AP link alone, omitting the UE-COW hop as a possible bottleneck; the claimed >2x gain may be inflated.","rationale":"The reader's ACCEPT is reasonable if the Aggressive relaying model is taken at face value. However, the central quantitative claim (Sec. I: more than two-fold SE increase) depends on the Aggressive branch, and the derivation of C†1 in Eq. (21) is internally inconsistent with the two-hop nature of the connection: when both hops reuse the same resources, the end-to-end SE is bounded by the weaker hop. The paper models the UE-COW hop in Sec. IV-A, including its blockage probability, but then discards that hop for the Aggressive strategy. Because the claimed >2x result is obtained in the dense-pedestrian regime where UE-COW blockage is most likely, this is load-bearing. This is not merely the acknowledged perfect-beam-alignment idealization; it is a structural modeling choice that can be tested by replacing C* with min(C⋆,C*). If the corrected Aggressive SE remains >2× Baseline, the paper's conclusion survives and the fix is a minor clarification. If it does not, the headline claim should be qualified to Conservative-only gains or to regimes with very low human density. Hence CONDITIONAL rather than REJECT: the issue is specific and checkable, and the paper's framework otherwise has independent support, including closed-form blockage expressions and simulation matches for the geometry relaxations.","tokens_in":9786,"tokens_out":6785,"duration_ms":70286,"concrete_test":"Recompute the Aggressive branch of Eqs. (21)–(23) with C†1(x0,xS)=min{C⋆(xS), C*(x1)} using the same parameters as Figs. 5 and 8; track P(C⋆<C*) over the integration region. If the Aggressive mean SE at 1 human/m² and 3–5 vehicles per 100 m falls below 2× Baseline, or if P(C⋆<C*) is non-negligible, the claimed gain is not supported and Eq. (21) needs revision. An alternative test: simulate the two-hop Aggressive relay with an actual decode-and-forward rate bottleneck (and optionally half-duplex constraints) and compare with the paper's analytical curve.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Section IV-C states that under the Aggressive strategy the joint UE-COW-AP connection is 'limited exclusively by the mean SE of the COW-AP link' and therefore sets C†1(x0,xS)=C*(x1) in Eq. (21). This is not a consequence of the system model. With the radio resources of the two hops overlapped, the end-to-end rate of a relayed flow cannot exceed the weaker of the two hops: C†1 should be min{C⋆(xS), C*(x1)} (or an explicit SINR formula if full-duplex relaying is assumed). The paper never establishes that C⋆(xS)≥C*(x1). In fact, because the COW antenna height hC is below hU (Sec. IV-A), the UE-COW link has a non-negligible human-blockage probability p⋆_B in Eq. (16); at the paper's dense-pedestrian operating point (1 human/m², Fig. 5), p⋆_B can be large, so C⋆ may be the bottleneck. Omitting this hop from Eq. (21) directly inflates the Aggressive curves in Figs. 5–8 and thus the headline 'more than two-fold increase' (Sec. I) for the Aggressive strategy. The Conservative branch correctly uses the harmonic combination in Eq. (7), so the asymmetry is internal. The simulation validation in Fig. 5 does not settle the point, since the simulator is described as relaxing geometry assumptions, not as implementing an independent two-hop rate model; it can reproduce the same modeling choice.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops an analytical framework for evaluating the spectral efficiency of mmWave vehicular relaying in a dense urban street deployment. It considers three connectivity strategies: a baseline with direct UE-AP links, a conservative relay strategy with orthogonal resource allocation for the two hops, and an aggressive relay strategy that allows overlapping resources. Using stochastic geometry, renewal theory, and 3GPP-compatible pathloss/blockage models, the authors derive closed-form expressions for the mean UE spectral efficiency and validate them via Monte Carlo simulation that relaxes several idealized geometry assumptions. The central quantitative claim is that aggressive vehicular relaying yields a more than two-fold increase in spectral efficiency in street deployments with dense pedestrian crowds and moderate vehicle density, with gains up to 120% depending on the fraction of vehicles acting as relays.","tokens_in":10045,"tokens_out":3887,"duration_ms":38191,"significance":"If correct, the framework would be a useful tool for 3GPP Rel-17 and beyond studies on mobile mmWave relaying, providing parameter-free, analytically tractable predictions. The paper is careful to derive blockage probabilities from first principles and to cross-check the analytical results with simulations that relax three major geometry assumptions. The parameters are taken from standards and prior measurement studies, with no free parameters fitted to the target output. The identification of regimes where relaying is most beneficial (medium vehicle density, high human density) is a concrete, falsifiable prediction. However, the validity of the central quantitative claim depends on a modeling step in the aggressive relay analysis that is questionable and is not independently validated by the simulation.","major_comments":[{"comment":"The Aggressive-relay joint SE is set to C†1(x0,xS)=C∗(x1), i.e., the COW-AP link alone, based on the assertion that the joint connection is 'limited exclusively by the mean SE of the COW-AP link.' This is not a consequence of the system model. For a relayed flow with overlapping radio resources, the end-to-end rate cannot exceed the weaker of the two hops, so the joint SE should be min{C⋆(xS), C∗(x1)} (or an explicit two-hop SINR expression), unless the paper proves that C⋆(xS)≥C∗(x1) for all relevant parameters. No such proof is provided. In fact, because the COW antenna height hC is below the UE height hU, the UE-COW link has a non-negligible human-blockage probability p⋆B given by Eq. (16); at the paper's dense-pedestrian operating point (1 human/m², Fig. 5) p⋆B can be large, so the UE-COW hop is a plausible bottleneck. The Conservative branch (Eq. (7)) correctly uses the harmonic combination of the two hop SEs, so the asymmetry is internal. The Monte Carlo validation in Fig. 5 does not resolve this point, because the simulation is described as relaxing geometry assumptions (pedestrian placement, vehicle centering, vehicle side blockage) rather than implementing an independent two-hop rate model; it can reproduce the same modeling choice. This omission directly inflates the Aggressive curves in Figs. 5–8 and the headline 'more than two-fold increase' claim in Section I. The authors should either derive the joint SE with the UE-COW hop explicitly included, or establish conditions under which the COW-AP link is always the bottleneck, and then revisit the numerical results and conclusions accordingly.","section":"IV-C, Eq. (21)"}],"minor_comments":[{"comment":"Both branches of Eq. (5) are written with the condition wS≤wU; the second branch should read wS>wU.","section":"III-C, Eq. (5)"},{"comment":"The simulation curves are shown without error bars or a statement of the number of Monte Carlo runs; a confidence interval would strengthen the validation claim.","section":"V, Fig. 5"},{"comment":"The text uses 'Agressive' (misspelled) in several places, including a subsection heading; it should be 'Aggressive'.","section":"II-C, IV-C"},{"comment":"The assumption that the UE 'instantaneously switches to the best available link via multi-connectivity mechanisms' is optimistic; a brief discussion of how finite switching latency and beam training overhead would affect the reported gains would be helpful.","section":"II-C"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the journal's scope and the analytical work is generally careful. The main concern is the Aggressive-relay modeling in Eq. (21), which is load-bearing for the headline claim; I would like the revision to address it directly. The manuscript also has a few presentation typos that should be fixed. No citation or attribution concerns."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things worth knowing up front. The paper gives a tractable renewal-theory framework for mobile mmWave relays in a street canyon, and the baseline and conservative-relay results look solid. But the aggressive-relay calculation contains a modeling error that inflates the headline \"more than two-fold\" gain; the stress-test note is right.\n\nThe genuinely new piece is the analytic mean-SE expression for the UE-COW-AP path with combined human/vehicle blockage, plus the conservative/aggressive resource-sharing comparison. That is a real extension of the authors' earlier blockage work, and it is aimed at a live 3GPP question (NR Rel. 17 mobile relays). The derivations are careful, the parameters are anchored to 3GPP TR 38.901, and the Monte Carlo check in Fig. 5 relaxes several geometric assumptions and matches. Credit where due: this is a competent, useful modeling effort.\n\nThe soft spot is not minor. In Eq. (21) the aggressive-strategy joint SE is set equal to C*(x1), the COW-AP link alone, on the grounds that the connection is \"limited exclusively\" by that link. That is an assumption, not a consequence. With overlapping resources, a relayed flow's end-to-end rate is bounded by the weaker of the two hops, so C†1 should be min{C⋆(xS), C*(x1)} (or an explicit full-duplex formula). The paper never shows C⋆≥C*, and the geometry points the other way in the dense-pedestrian regime: hC<hU, and p⋆_B in Eq. (16) is substantial at 1 human/m². So the aggressive curves in Figs. 5–8 are inflated, and the >2x claim rests on that. The simulation in Fig. 5 does not rescue it: it is described as relaxing geometry, not implementing an independent two-hop rate model, so it can reproduce the same choice. The conservative branch uses the harmonic combination, so the asymmetry is internal.\n\nMinor items: Eq. (5) repeats wS≤wU in both branches; no error bars in Fig. 5; perfect beam alignment and instantaneous switching are explicit but standard first-order assumptions.\n\nWho should spend time on this? People working on mmWave relay simulation or 3GPP evaluations. The qualitative conclusions (dense pedestrians plus moderate traffic is the high-gain regime; 20–40% COWs captures most of the benefit) likely survive for conservative relaying, but the aggressive numbers need rework or heavy caveats. It deserves a serious referee, not a desk reject. The right review would ask the authors to correct the hop bottleneck and either verify C⋆≥C* numerically or present the min expression.","headline":"Solid baseline and conservative relaying framework, but a load-bearing hop-bottleneck error in the aggressive strategy inflates the headline gain.","tokens_in":10631,"tokens_out":3401,"would_cite":false,"duration_ms":34384,"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 develops an analytical framework showing that vehicular mmWave relaying can more than double mean spectral efficiency in dense urban street deployments when users can select the best link.","keywords":["millimeter-wave relaying","vehicular relays","spectral efficiency","urban street deployment","human-body blockage","vehicle-body blockage","multi-connectivity","5G New Radio"],"falsifier":"Run a 28 GHz street-deployment field trial or detailed simulator that keeps the same geometry and traffic densities but models finite beam training, imperfect alignment, and handover latency; if relay-aided mean spectral efficiency no longer exceeds the direct-to-base baseline by roughly a factor of two for the Aggressive strategy, the paper's headline quantitative claim would be refuted.","tokens_in":1664,"feed_emoji":"📶","tokens_out":2434,"duration_ms":87582,"temperature":0.7,"pith_summary":"An analytical model of a single urban street segment shows that car-mounted millimeter-wave relays can more than double the mean spectral efficiency of a 5G user, from about 8-12 bits/s/Hz to 17-18 bits/s/Hz, provided pedestrians are dense and vehicle traffic is moderate. The model accounts for blockages by human bodies and by cars and buses, and compares a direct-to-base-station baseline with two relay strategies: Conservative, which uses separate radio resources for the user-to-relay and relay-to-base hops, and Aggressive, which lets those hops reuse resources. The Aggressive strategy yields a 70-120% improvement in mean spectral efficiency, while the Conservative strategy yields only 8-12%. The gains grow with pedestrian density, peak at medium vehicle densities around 3-5 vehicles per 100 m, and stop growing once about 40% of vehicles act as relays. If correct, these results give operators a quantitative case for deploying vehicular relays selectively in blockage-heavy streets rather than everywhere.","feed_headline":"Vehicular relays can double urban mmWave 5G efficiency","feed_subtitle":"Car-mounted relays add 70-120% spectral efficiency in dense pedestrian streets, peaking at moderate traffic.","key_machinery":"The machinery is a sequence of blockage-probability formulas for the three links in the system, user to base station, user to relay car, and relay car to base station, computed with renewal theory over random pedestrian spacings and random car or bus spacings. Human-body blockage is treated as a pedestrian path crossing a rectangular blockage zone; vehicle blockage depends on whether a bus is tall enough to occlude the line of sight, with a critical height that turns out not to depend on user position. These probabilities feed mean spectral efficiency expressions built from 3GPP pathloss and directional antenna gains; for the two-hop relay path, the Aggressive strategy uses the relay-to-base link's efficiency directly, while the Conservative strategy combines the two hop efficiencies harmonically. The user's final efficiency is the maximum of the direct and relayed values, and the paper verifies the closed-form analysis against a simulation that relaxes three geometric assumptions.","core_discovery":"The paper's central claim is that intelligent users who continuously select the strongest available link, direct to a static mmWave access point or through a relay-equipped car, can more than double their mean spectral efficiency in a street canyon with dense human crowds and moderate vehicular traffic. The quantitative support is a comparison of three connectivity strategies: a baseline that ignores relays, a Conservative relay strategy that avoids any resource overlap between hops, and an Aggressive relay strategy that reuses radio resources across hops. In the modeled 28 GHz deployment with base stations 300 m apart, the Aggressive strategy raises mean UE spectral efficiency to about 17-18 bits/s/Hz, a 70-120% improvement over baseline, while the Conservative strategy manages only an 8-12% gain. The paper further identifies the conditions where relaying pays off most: dense pedestrian sidewalks, a medium density of vehicles, and relay coverage ranges large enough that a relay is nearly always present, after which further increases in relay fraction or coverage range add little.","pith_inferences":["Because the model assumes perfect beam alignment and instantaneous link switching, the practical gain in a deployed system will be lower; the next step is to add beam-training time, misalignment statistics, and handover latency to the same analysis and see how much of the two-fold gain survives.","The renewal-theoretic blockage analysis could be extended to moving relays by treating a relay's position relative to a user as time-varying, allowing estimates of how often the best-link switch occurs and how long a user stays in a relay's coverage.","The saturation near 40% relay penetration suggests an economically useful design rule: recruit only a bounded share of vehicles as relays, which limits incentives, energy budgets, and radio-resource overhead for participating vehicles.","The Aggressive strategy's gain rests on spatial isolation of narrow beams; a testable extension is to replace the perfect-alignment assumption with actual beam-sweeping protocols and verify the two-fold gain in an interference-aware system-level simulation."],"forward_implications":["Deploying vehicular relays pays off mainly in pedestrian-dense street canyons, where human-body blockage is the dominant impairment of direct mmWave links.","Relay gains are non-monotonic in vehicle density: too few vehicles means too few relays, while too many vehicles introduces vehicle-body blockage, with the best returns at roughly 3-5 vehicles per 100 m.","Equipping more than about 40% of vehicles as relays yields little additional mean spectral efficiency, so network operators can cap relay participation without losing most of the benefit.","Even a 10% fraction of relay-equipped vehicles captures most of the Aggressive-strategy gain, so a small fleet of specially equipped vehicles could provide most of the improvement.","The choice between Conservative and Aggressive relaying changes the outcome substantially: resource reuse across hops is what unlocks the two-fold gain, while disjoint resource allocation limits gains to under 15%."],"supporting_citations":[{"why":"Supplies the 3GPP-compatible pathloss model and node heights used in every SNR and spectral-efficiency calculation.","marker":"[9]"},{"why":"Establishes the blockage-zone method for deciding when a pedestrian blocks a mmWave link.","marker":"[13]"},{"why":"Provides the renewal-theory result used to compute the probability that a pedestrian or vehicle occupies the blockage zone.","marker":"[14]"},{"why":"Provides the renewal-process result used for vehicle-body blockage probability on the street.","marker":"[15]"},{"why":"Justifies the assumption that a user instantly switches to the best available link via multi-connectivity.","marker":"[12]"},{"why":"Supports the Aggressive strategy's premise that narrow-beam mmWave links can reuse radio resources with limited interference.","marker":"[11]"},{"why":"Supplies the numerical technique for computing the CDF of the distance between a relay car and the nearest bus.","marker":"[16]"},{"why":"Provides the UE transmit power and antenna-gain values used in the numerical results.","marker":"[17]"}],"fun_headline_variants":["Smart users double mmWave speed with car relays","Vehicular relaying lifts urban mmWave efficiency by up to 120%","Car relays double mmWave 5G throughput in crowded streets","Aggressive vehicular relaying yields 120% mmWave gain","Urban mmWave 5G doubles with smart vehicular relays"],"cache_read_input_tokens":12672,"weakest_assumption_plain":"The central premise is that every communicating pair has ideally aligned beams and that a user can switch instantly to whichever link, direct or relayed, is strongest, so the reported gains assume away beam-training time, misalignment, and handover latency.","fun_headline_variants_meta":{"raw":{"variants":["Smart users double mmWave speed with car relays","Vehicular relaying lifts urban mmWave efficiency by up to 120%","Car relays double mmWave 5G throughput in crowded streets","Aggressive vehicular relaying yields 120% mmWave gain","Urban mmWave 5G doubles with smart vehicular relays"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000457,"raw_usage":{"total_tokens":2285,"prompt_tokens":928,"completion_tokens":1357,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":544,"completion_tokens_details":{"reasoning_tokens":1269}},"tokens_in":544,"tokens_out":1357,"duration_ms":9891,"temperature":1.0,"reasoning_tokens":1269,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:00:01.690750+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a 28 GHz street-deployment field trial or detailed simulator that keeps the same geometry and traffic densities but models finite beam training, imperfect alignment, and handover latency; if relay-aided mean spectral efficiency no longer exceeds the direct-to-base baseline by roughly a factor of two for the Aggressive strategy, the paper's headline quantitative claim would be refuted.","supporting_citations":[{"cited_title":"Study on channel model for frequencies from 0.5 to 100 GHz (Release 14),","cited_arxiv_id":null,"evidence_quote":"Supplies the 3GPP-compatible pathloss model and node heights used in every SNR and spectral-efficiency calculation."},{"cited_title":"Dynamic multi-connectivity performance in ultra- dense urban mmWave deployments,","cited_arxiv_id":null,"evidence_quote":"Establishes the blockage-zone method for deciding when a pedestrian blocks a mmWave link."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the renewal-theory result used to compute the probability that a pedestrian or vehicle occupies the blockage zone."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the renewal-process result used for vehicle-body blockage probability on the street."},{"cited_title":"An efficient uplink multi-connectivity scheme for 5G millimeter-wave control plane applications,","cited_arxiv_id":null,"evidence_quote":"Justifies the assumption that a user instantly switches to the best available link via multi-connectivity."},{"cited_title":"Interference and SINR in millimeter wave and terahertz communication systems with blocking and directional antennas,","cited_arxiv_id":null,"evidence_quote":"Supports the Aggressive strategy's premise that narrow-beam mmWave links can reuse radio resources with limited interference."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the numerical technique for computing the CDF of the distance between a relay car and the nearest bus."},{"cited_title":"5G cellular User Equipment: From theory to practical hardware design,","cited_arxiv_id":null,"evidence_quote":"Provides the UE transmit power and antenna-gain values used in the numerical results."}],"review_version":1}