{"id":"bb5574c8-f125-48db-a325-544925daa757","arxiv_id":"1908.06842","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper presents closed-form packet error probability and a Stackelberg power-pricing solution for cooperative V2V/V2I uplinks with correlated RSU antennas, but the EC formula and game optimality proof are flawed.","lead":"This paper derives closed-form expressions for packet error probability in cooperative vehicle-to-vehicle and vehicle-to-infrastructure uplinks when the roadside unit's antennas experience correlated fading, and a Stackelberg game to set transmit power and helper pricing. It targets system designers who need reliability formulas under realistic antenna correlation, but several central derivations contain errors.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proposition 1's optimal power-split formula is invalid: the proof maximizes only the sigmoid SNR term and drops the source's payment term, so the claimed 'if and only if' condition fails.","rationale":"The paper's central claims include both the packet-error analysis and the Stackelberg optimality results. The reader's verdict of REJECT is supported, but my chosen load-bearing concern is in the game-theoretic section rather than the EC distribution. The Proposition 1 proof is a clear internal inconsistency: it omits the −p_i(1−φ)P cost term when maximizing U_s. The same omission propagates into Proposition 2 and the derived P* and p_i* expressions, so the claimed equilibrium characterization is not established. This is a definite mathematical flaw rather than an assumption that might be verified by simulation. The EC Gamma formula may well be correct as a standard Kotz–Adams result, and the reader's concern there is better framed as a verification request than as a confirmed error; this is why my agreement with the reader's stated weakest_assumption is only partial. Even without relying on the EC issue, the invalid optimality proofs are sufficient to reject the paper in its current form, since the abstract explicitly promises optimal transmit-power and pricing solutions. The proposed numerical maximization test would settle the Proposition 1 concern directly and unambiguously.","tokens_in":12341,"tokens_out":14439,"duration_ms":156633,"concrete_test":"Fix one channel realization and parameters, and numerically maximize U_s(φ)=w_p/[1+exp(−a(min{Aφ,B(1−φ)}−γ0))]−p_i(1−φ)P over φ∈[0,1], where A=P|h|²/(d_{VsVi}^α N0) and B=Pη/(d_{Vi*R}^α N0). Compare the numerical argmax with Eq. (23). Choosing p_iP > aB w_p/4 makes the payment term dominate near the SNR-balancing point, and the argmax will exceed Eq. (23), directly falsifying Proposition 1.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The game-theoretic contribution rests on Proposition 1, which claims that U_s is maximized if and only if φ equals η d^α_{VsVi}/(η d^α_{VsVi}+|h_{VsVi}|^2 d^α_{Vi*R}). The proof maximizes only the sigmoid satisfaction term U_R, but the actual utility is U_s = w_p U_R − p_i(1−φ)P. The payment term is strictly decreasing in φ, so the source can gain by choosing φ above the SNR-balancing point, or even by setting φ=1 and paying nothing while accepting U_s≈0. Thus Eq. (23) maximizes the SNR component, not the total utility in Eq. (21). Proposition 2 is also internally inconsistent: the second derivative in Eq. (26) reduces to c^2 w_p f(1−f)(1−2f) with c=aη|h|²/ϖ, which is positive whenever f<1/2, i.e. for sufficiently low P; this contradicts the claim that ∂²U_s/∂P² is always negative. Consequently the closed-form optimal power in Eq. (27) and the optimal price in Eq. (29), which are derived from these claims, are unsupported. The EC Gamma assumption raised by the reader is less decisive: Eq. (16) is the standard Kotz–Adams result for a specific exponential-correlation gamma model, although a direct Monte Carlo check would still be prudent.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies an uplink cooperative V2V/V2I network in which a source vehicle selects one of several helper vehicles and the selected helper forwards the packet to a multi-antenna RSU over Nakagami-m faded links. The first contribution is a packet-error analysis for two antenna-correlation models at the RSU, constant correlation (CC) and exponential correlation (EC), combining order statistics for the best helper with MRC at the RSU. The second contribution is a Stackelberg game in which the source chooses the power split phi and total power P, and the selected helper sets a per-watt price, with closed-form expressions for the optimal phi, P, and price. The authors compare their analytical packet-error expressions with MATLAB Monte Carlo simulations in Figures 2-4.","tokens_in":12664,"tokens_out":11551,"duration_ms":125361,"significance":"If the derivations were correct, the closed-form packet-error expressions under correlated RSU antennas would be a useful addition to the vehicular-communications literature, and the Stackelberg formulation is a reasonable and relevant framework for cooperative V2X. The Monte Carlo verification reported in Figures 2-4 and the absence of fitted parameters are strengths of the presentation. However, the printed mathematical derivations contain load-bearing errors in both main contributions: the EC CDF in Eq. (18) is not a valid CDF, the Gamma CDF in Eq. (11) omits the Nakagami shape parameter inside the incomplete-Gamma argument, and the game-theoretic optimization in Proposition 1 maximizes the SNR-satisfaction term while dropping the payment term from the source's utility. These are not presentation issues; they invalidate the central analytical results as written.","major_comments":[{"comment":"The EC cumulative distribution function is invalid as written. Integrating the Gamma density in Eq. (16) from 0 to gamma0 yields the regularized lower incomplete Gamma function gamma(mM^2/lambda, Mm*gamma0/(lambda*gamma_bar))/Gamma(mM^2/lambda), not the upper incomplete Gamma function. Since Eq. (18) explicitly identifies Gamma(.,.) as the upper incomplete Gamma function, the expression tends to Gamma(mM^2/lambda) > 0 as gamma0 approaches 0 and is not bounded by 1 for small thresholds. This invalidates the EC packet-error probability obtained from Eqs. (7)-(19) and the EC results in Figures 2-4, even if one grants the Kotz-Adams distributional assumption in Eq. (16).","section":"Section III, Eq. (18)"},{"comment":"The Nakagami-m SNR CDF is written incorrectly. From the PDF in Eq. (10), the probability that gammaVsVi is below gamma0 is gamma(m, m*gamma0/gamma_barVsVi)/Gamma(m), but Eq. (11) omits the factor m inside the lower incomplete Gamma function. This factor is essential for all m > 1 and even affects the Rayleigh case if the same formula is used. Since Eq. (11) is multiplied into the packet-error expression in Eq. (7) for both the CC and EC models, every packet-error probability derived from the printed equations is affected.","section":"Section III, Eq. (11)"},{"comment":"The proof of Proposition 1 maximizes U_R, not U_s. The source utility in Eq. (21) is U_s = w_p U_R - p_i(1-phi)P, and the payment term -p_i(1-phi)P is strictly increasing in phi. Therefore the SNR-balancing point gammaVsVi* = gammaVi*R does not in general maximize U_s; for example, phi = 1 gives zero payment and utility w_p/(1+exp(a*gamma0)), which can exceed the value at Eq. (23) when p_i is large. The claimed 'if and only if' condition is false, and the rest of the game-theoretic analysis, including Eq. (24), the optimal power in Eq. (27), and the optimal price in Eq. (29), relies on this incorrect optimizer.","section":"Section IV, Proposition 1 and Eq. (23)"},{"comment":"The claim that the second derivative in Eq. (26) is always negative is false. Let f = sigma(a(eta P |hVsVi|^2/varpi - gamma0)). Then d^2 U_s/dP^2 reduces to a^2 (eta |hVsVi|^2/varpi)^2 w_p f(1-f)(1-2f). This is positive whenever the instantaneous SNR is below the threshold gamma0, i.e., whenever f < 1/2. Consequently, the concavity argument, the first-order optimality condition, and the closed-form optimal power in Eq. (27) are not supported by the provided derivation.","section":"Section IV, Proposition 2 and Eq. (26)"},{"comment":"The finite sum in the CC CDF appears to run from n = 0 to (1-rho_c+M*rho_c)/(M*rho_c) - 1. For the parameters used in Section V (e.g., M = 10, rho_c = 0.1), this upper limit is 0.9, which is not an integer. Since no integrality condition is stated, the expression in Eq. (15) is undefined for the simulated parameter regime, and the CC packet-error curves do not follow from the printed formula.","section":"Section III, Eq. (15)"}],"minor_comments":[{"comment":"The exponential factor is written as exp(-m s / zbar) with an undefined variable s; it should be exp(-m z / zbar) to match the Gamma density.","section":"Section III, Eq. (10)"},{"comment":"The proof states '0 < hVsVi, wp < 1' as if hVsVi were a probability. The fading coefficient hVsVi is generally complex or a real envelope, and only its magnitude enters the SNR; the concavity condition should be phrased in terms of |hVsVi| > 0.","section":"Section IV, Proposition 2"},{"comment":"The axis label 'rho_c = rho_c' appears to be a typo; it should read 'rho_c = rho_e'.","section":"Section V, Figure 4"},{"comment":"The placement of parentheses in the finite-sum expression in Eq. (15) is ambiguous as typeset, and the derivation of the CC CDF would be much easier to verify if the citation to the specific identity in [27] were given with equation numbers.","section":"Section III, Eqs. (14)-(15)"},{"comment":"The helper's optimal price in Eq. (29) contains no dependence on w_p, a, or P, which is surprising for a Stackelberg equilibrium. Even setting aside the invalid Proposition 1, this suggests that the derivation should be re-examined and the result re-derived from the true first-order condition.","section":"Section IV, Eq. (29)"}],"recommendation":"reject","confidential_remarks":"The manuscript header states that it is 'ACCEPTED FOR IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS.' The load-bearing mathematical errors in the EC CDF, the Nakagami CDF, and the game-theoretic optimization are substantive and would affect any published version containing these equations. I would advise the editor to verify whether the accepted/published version contains the same errors and, if so, to consider a correction or withdrawal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nQuick take: this paper applies known correlated-Gamma results to a cooperative V2V/V2I uplink and gets closed-form packet error probabilities, plus a Stackelberg power/price game. The application is legitimately new in context, and the packet-error part is mostly standard. But there are load-bearing errors in both the EC formula and the game proofs, so it reads as an accepted-but-not-correct paper.\n\nWhat's good: using Gurland's constant-correlation and Kotz-Adams exponential-correlation distributions for the RSU combiner output is a sensible extension, and the authors do verify the intended formulas with MATLAB (though the printed EC expression is wrong, see below). The writing is clear and the literature coverage is adequate.\n\nSoft spots: (18) gives the EC CDF as Γ(mM²/λ, ...), which they define as the upper incomplete gamma function. That is not a CDF: at γ0=0 it equals Γ(a), not 0, and it decreases to 0. The correct expression is the regularized lower incomplete gamma. This is not a typo in a minor equation; it is the main closed-form result for EC. Either the formula in the paper is wrong, or the simulations used a different expression.\n\nSecond, Proposition 1 claims U_s is maximized iff φ balances the two hop SNRs, but the proof maximizes only the sigmoid satisfaction term U_R and ignores the payment term -p_i(1-φ)P. Since that cost term is decreasing in φ, the source can prefer a larger φ (less money paid to the helper) even if it lowers the SNR slightly. The 'if and only if' fails. Proposition 2's claim that ∂²U_s/∂P² < 0 is also false: for arguments where the sigmoid is below 1/2 (low SNR), the second derivative is positive. Thus the concavity argument and the closed-form optimal power are unsupported.\n\nThe EC Gamma-sum assumption (16) is actually the standard Kotz-Adams result, so that specific piece is not the issue. The problems are the printed CDF and the game-theoretic optimization.\n\nNet: the packet-error analysis is salvageable with a corrected (18) and a fresh look at the simulation match; the game section needs real rework. This paper deserves serious refereeing because the modeling idea is relevant and the errors are checkable, but it should not be accepted as is. I wouldn't cite it in its current form, and I wouldn't spend a reading group on it unless the goal is to illustrate how utility functions get mis-optimized.\n\nRecommendation: send it to reviewers, but expect heavy revision or rejection.","headline":"A useful but flawed extension: the EC CDF is misprinted and the Stackelberg proofs drop key terms, so the paper is not publishable as is.","tokens_in":13163,"tokens_out":4587,"would_cite":false,"duration_ms":44336,"reading_group":"no","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper derives closed-form packet-error expressions for cooperative V2V/V2I uplinks with correlated antennas at the roadside unit, and solves a Stackelberg game for optimal transmit power and helper price.","keywords":["Antenna correlation","Stackelberg game","Vehicle-to-infrastructure (V2I)","Vehicle-to-vehicle (V2V)","Packet error probability","Nakagami-m fading","Cooperative vehicular networks","Maximum ratio combining"],"falsifier":"Simulate the full two-hop model: draw Nakagami-m fading for the first hop, select the best helper, draw exponentially or constantly correlated Nakagami-m coefficients for the roadside-unit antennas according to $\\rho_e$ or $\\rho_c$, apply maximum-ratio combining, and compare the empirical packet error with equations (15), (18), and (19) over a grid of $m$, $M$, $N$, and correlation values; a systematic mismatch, especially in the exponential-correlation case, would falsify the closed-form claim.","tokens_in":12176,"feed_emoji":"📡","tokens_out":11356,"duration_ms":109530,"temperature":0.7,"pith_summary":"The paper aims to show that the packet error probability of a cooperative vehicle-to-vehicle/vehicle-to-infrastructure uplink can be computed in closed form even when the roadside unit's closely packed antennas are not independent. Two correlation models are treated: constant correlation, where every antenna pair shares the same correlation, and exponential correlation, where correlation decays with antenna spacing. The derivation covers Nakagami-m fading, a flexible model that reduces to Rayleigh fading when m equals one, best-helper selection on the first hop, and maximum-ratio combining at the roadside unit on the second hop. It also solves a Stackelberg game in which the source vehicle buys forwarding power from a helper at a per-Watt price, yielding explicit formulas for optimal transmit power and asking price. If correct, these formulas let designers evaluate reliability, diversity gains, and pricing behavior directly from the fading parameter, correlation coefficients, number of helpers, and number of antennas.","feed_headline":"Closed-form packet-error curves for cooperative V2V links","feed_subtitle":"New equations cover correlated roadside-unit antennas, fading, and helper count—plus a game for power and price.","key_machinery":"The load-bearing object is the bottleneck relationship $\\gamma_{e2e} = \\min\\{\\gamma_{V_s V_{i^*}}, \\gamma_{V_{i^*} R}\\}$ between the best first-hop SNR and the maximum-ratio-combined second-hop SNR at the roadside unit. On the first hop, order statistics turn the best-helper outage into a product of Gamma cumulative distribution functions; on the second hop, the correlated antenna sum is represented by two densities: a confluent-hypergeometric density for constant correlation and a Gamma density for exponential correlation with shape $mM^2/\\lambda$ and scale $\\lambda \\bar{\\gamma}/(Mm)$. The Stackelberg machinery adds a sigmoid satisfaction function for the source, a per-Watt price for the helper, and a leader-follower equilibrium that determines the power-split factor $\\phi$, the transmit power $P$, and the price $p_i$.","core_discovery":"The paper's central claim is that for a two-phase decode-and-forward uplink, the end-to-end packet error probability under correlated roadside-unit antennas is available in closed form: one expression for constant correlation involving hypergeometric functions, and a compact expression for exponential correlation involving an upper incomplete Gamma function. The first-hop outage, after selecting the helper with the largest SNR, becomes a product of incomplete Gamma functions, equation (11). The second-hop outage is obtained by integrating the combined SNR density at the maximum-ratio combining output, using the constant-correlation density (13) or the exponential-correlation Gamma-sum density (16), yielding equations (15) and (18). Combining the two hops through the bottleneck SNR and compounding over L blocks gives the total packet error probability in equation (19). On the game side, the paper proposes a Stackelberg game with the source as leader and helpers as followers, proves concavity of the source utility in transmit power, and derives the optimal transmit power (27) and helper asking price (29).","pith_inferences":["The exponential-correlation formula can be tested separately from the rest of the paper: generate exponentially correlated Nakagami-$m$ samples, apply maximum-ratio combining, and compare empirical outage with equation (18) across a grid of $\\rho_e$, $m$, and $M$; this would isolate whether the Gamma-sum approximation is the limiting step.","Because the closed forms are algebraic in $m$, $\\rho$, $M$, and $N$, they could be embedded in a system-level design loop that sweeps roadside-unit antenna spacing or helper recruitment thresholds, something the paper itself does not demonstrate.","Inserting imperfect channel estimation into the helper-selection and game stages would shift both the selected helper and the equilibrium power and price; the paper identifies this as future work, and a natural quantitative extension is to measure how sensitive equations (27) and (29) are to such errors."],"forward_implications":["More helper vehicles reduce packet error probability through selection diversity on the first hop, with diminishing returns as the distance ratio $d_{V_s V_{i^*}}/d_{V_{i^*} R}$ grows large.","Higher antenna correlation at the roadside unit increases packet error, and the constant- and exponential-correlation predictions converge as the correlation coefficients approach one.","Packet error worsens as the SNR threshold $\\gamma_0$ rises and as a packet is split into more blocks, so higher vehicle speeds shorten coherence time and reduce reliability.","The Stackelberg solution gives a finite optimal transmit power that balances the source's satisfaction against the helper's price, and an asking price that maximizes the helper's profit.","Optimal transmit power increases with both source-to-helper and helper-to-roadside-unit distances, and decreases when the Nakagami parameter $m$ grows, for both correlation models."],"supporting_citations":[{"why":"It supplies the probability density of the combined SNR at the maximum-ratio combining output under constant correlation, which is the basis of equations (13) through (15).","marker":"[26]"},{"why":"It supplies the Gamma distribution for the sum of exponentially correlated Gamma variables, which is the load-bearing assumption behind equation (16) and the exponential-correlation closed form (18).","marker":"[28]"},{"why":"It gives the integral identities used to resolve the constant- and exponential-correlation integrals into the closed-form expressions.","marker":"[27]"},{"why":"It provides the Gamma density for Nakagami-m SNR that underlies the first-hop outage expression in equation (11).","marker":"[25]"},{"why":"It justifies treating packet error probability as an outage-based lower bound under ideal coding, which connects the block-fading model to the final packet error expression.","marker":"[24]"},{"why":"It provides the Stackelberg game-theoretic framework that the paper applies to optimize transmit power and helper price.","marker":"[29]"}],"fun_headline_variants":["Correlated antennas? New exact error equations for V2V","Closed-form V2V error under correlated fading","Game-theoretic pricing for cooperative V2V links","Exact packet-error probabilities for correlated V2V uplinks","Cooperative V2V with correlated antennas: exact error curves"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the sum of exponentially correlated Nakagami-m antenna signals at the roadside unit is exactly the Gamma distribution written in equation (16); if that borrowed distributional shortcut diverges from the true correlated sum, the exponential-correlation packet-error formulas do not follow.","fun_headline_variants_meta":{"raw":{"variants":["Correlated antennas? New exact error equations for V2V","Closed-form V2V error under correlated fading","Game-theoretic pricing for cooperative V2V links","Exact packet-error probabilities for correlated V2V uplinks","Cooperative V2V with correlated antennas: exact error curves"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001203,"raw_usage":{"total_tokens":4979,"prompt_tokens":987,"completion_tokens":3992,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":603,"completion_tokens_details":{"reasoning_tokens":3911}},"tokens_in":603,"tokens_out":3992,"duration_ms":30711,"temperature":1.0,"reasoning_tokens":3911,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:59:32.220415+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate the full two-hop model: draw Nakagami-m fading for the first hop, select the best helper, draw exponentially or constantly correlated Nakagami-m coefficients for the roadside-unit antennas according to $\\rho_e$ or $\\rho_c$, apply maximum-ratio combining, and compare the empirical packet error with equations (15), (18), and (19) over a grid of $m$, $M$, $N$, and correlation values; a systematic mismatch, especially in the exponential-correlation case, would falsify the closed-form claim.","supporting_citations":[{"cited_title":"Distribution of the maximum of the arithmetic mean of correlated random variables,","cited_arxiv_id":null,"evidence_quote":"It supplies the probability density of the combined SNR at the maximum-ratio combining output under constant correlation, which is the basis of equations (13) through (15)."},{"cited_title":"Distribution of sum of identically distributed exponentially correlated gamma-variables,","cited_arxiv_id":null,"evidence_quote":"It supplies the Gamma distribution for the sum of exponentially correlated Gamma variables, which is the load-bearing assumption behind equation (16) and the exponential-correlation closed form (18)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It gives the integral identities used to resolve the constant- and exponential-correlation integrals into the closed-form expressions."},{"cited_title":"Performance modeling and analysis on conditional DF relaying scheme over Nakagami-m fading channels with integral m,","cited_arxiv_id":null,"evidence_quote":"It provides the Gamma density for Nakagami-m SNR that underlies the first-hop outage expression in equation (11)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It justifies treating packet error probability as an outage-based lower bound under ideal coding, which connects the block-fading model to the final packet error expression."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the Stackelberg game-theoretic framework that the paper applies to optimize transmit power and helper price."}],"review_version":1}