{"id":"209ffdab-1ec1-4584-8eac-5dbb6819c5f4","arxiv_id":"2411.16441","paper_version":2,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper claims exact one-turn shortest-path length distributions in Poisson line Cox processes, plus a two-turn upper bound, but the derivations contain internal inconsistencies and likely factor errors.","lead":"This paper derives closed-form formulas for the distribution of the shortest street path to the nearest point in a random network of lines with random points on them, under a restriction of at most one turn. It also gives bounds for two-turn paths and sketches applications to vehicle communication and electric-vehicle charging placement.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1's stated CCDF contradicts its own proof: the proof yields exp(-2μt-4λt+(2λ/μ)(1-e^{-2μt})), while Eq. (1) halves the λ terms, and the two-sided intersection PPP has mean 4λt, leaving the central exact formula unsupported.","rationale":"The reader's weakest assumption identifies the same load-bearing arithmetic in Theorem 1: the intersection count on a two-sided segment should have mean 4λt, not 2λt. My stress-test sharpens this to a direct contradiction between the theorem statement and its proof: the proof's displayed computation yields an exponent with -4λt + (2λ/μ)(1-e^{-2μt}), while Eq. (1) says -2λt + (λ/μ)(1-e^{-2μt}). This is not a matter of interpretation or normalization; the theorem as stated cannot be derived from the proof as written. The error is load-bearing because Eq. (1) is the first main result, is used in the numerical comparisons, and feeds directly into the RIS application in Eq. (18). The proposed Monte Carlo check would settle which expression is correct, but regardless of the outcome, the paper's exact formula and its applications are not supported without substantial correction. Theorem 2 shows additional internal inconsistencies between its piecewise Z and the proof's Z1 formulas, reinforcing the verdict. I therefore agree with the reader's REJECT and see no reason to adjust the verdict.","tokens_in":78,"tokens_out":9412,"duration_ms":206052,"concrete_test":"Recompute the one-turn CCDF from the typical point by direct integration over the two-sided intersection PPP: P(D>t) = e^{-2μt} · E[∏_{i: S_i ∈ PPP(2λ) on [-t,t]} e^{-2μ(t-|S_i|)}] = exp(-2μt - ∫_{-t}^{t} 2λ(1-e^{-2μ(t-|s|)}) ds). Evaluate at λ=μ=1, t=1: this yields exp(-2-4+2(1-e^{-2})) ≈ 0.0139, whereas Eq. (1) gives exp(-2-2+1-e^{-2}) ≈ 0.0433. Then run a Monte Carlo simulation of the PLCP (sample PLP lines crossing the segment [-1,1], place PPP points of intensity 1 on each line, measure the empirical CCDF at t=1) and compare with both expressions. If the simulation matches the corrected integration and not Eq. (1), the theorem statement is false and all downstream applications need re-derivation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is the exact one-turn shortest-path CCDF, and Theorem 1 (Eq. (1)) is used directly in the applications of Section VI. However, Theorem 1 as stated is internally inconsistent with its own proof. In the proof, after conditioning on k intersections, the authors write pk(t)=e^{-2λt}(2λt)^k/k! and average e^{-2μ(t-s)} over s∈[0,t], which evaluates to exp(-2μt - 4λt + (2λ/μ)(1-e^{-2μt})). The theorem statement instead gives exp(-2μt - 2λt + (λ/μ)(1-e^{-2μt})), exactly halving the λ-dependent terms. Moreover, the proof's own model is also inconsistent with the paper's definitions: the typical point lies on Lx, and the one-turn path may start in either direction, so the relevant intersections are on the two-sided segment [-t,t], which under Definition 1 forms a PPP of intensity 2λ per unit length with mean 4λt, not 2λt. The factor exp(-2μt) for the void event on Lx correctly uses this two-sided segment of length 2t, so the line-intersection count must be defined on the same segment. Integrating over the two-sided intersection PPP gives exp(-∫_{-t}^{t} 2λ(1-e^{-2μ(t-|s|)})ds) = exp(-4λt + (2λ/μ)(1-e^{-2μt})). Regardless of which normalization is intended, Eq. (1) does not follow from the proof; the displayed expression in the proof and the theorem statement differ by a factor of two in the λ terms. Since Eq. (1) is subsequently used to compute FD in the RIS and charging applications, the quantitative claims inherit this error. The same type of discrepancy appears in Theorem 2, where the piecewise Z expression in the statement does not match the Z1 formulas in the proof (e.g., the first case has (sinω - sinω1)/sin(ω1-ω) instead of the proof's (sinθ + sinθ1)/sin(θ1-θ)). Thus the main exact results are not supported as written.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies shortest path (ℓ1) distances in Poisson line Cox processes (PLCPs), where points of intensity μ sit on the lines of a Poisson line process (PLP) of density λ. Under a one-turn restriction, it claims exact closed-form complementary cumulative distribution functions (CCDFs) for the nearest-neighbor path length from the typical PLCP point (Theorem 1) and from the typical PLP intersection (Theorem 2), plus an upper bound for the two-turn case from the typical point (Theorem 3). The results are then applied to RIS-assisted V2V communication and electric-vehicle charging placement. The central claims are presented as exact, parameter-free characterizations of a quantity whose distribution has been open.","tokens_in":12167,"tokens_out":9578,"duration_ms":79329,"significance":"If the claimed results were correct, they would constitute a substantial contribution to stochastic geometry: exact path-length distributions for an isotropic Poisson line Cox process have been open, and the applications to vehicular networks and infrastructure planning are relevant and well motivated. The paper is self-contained in its derivations, does not fit parameters to match simulations, and explicitly identifies the gap in prior work. However, the central theorems contain internal inconsistencies with their own proofs and with the model definitions, and these inconsistencies invalidate the quantitative statements that are subsequently used in the applications. The paper therefore cannot be accepted in its present form.","major_comments":[{"comment":"The proof of Theorem 3 contains a Poisson-parameter mismatch. After stating that the number n1 of lines intersecting Lx in [0,s1] is Poisson with parameter 2s1λ, the expectation is written with exp(-2λs1)(λs1)^k/k! instead of exp(-2λs1)(2λs1)^k/k!. Consequently the displayed simplification to exp(-λ∫[2 - T(w,u) f(w)] dw) corresponds to a Poisson mean of λs1, not 2λs1. This affects the bound in Eq. (14) and the subsequent numerical evaluation in Fig. 8, even if the geometric construction of the event Ei is accepted.","section":"Section IV, Theorem 3"}],"minor_comments":[{"comment":"Remark 1 states that exp(-4λt) corresponds to the event that no line intersects within distance t, but Theorem 1 uses -2λt in the exponent; resolve this inconsistency and align the remark with the corrected theorem.","section":"Remark 1 vs Theorem 1"}],"recommendation":"reject","confidential_remarks":"The paper's central contributions are exact one-turn path-length distributions, but the two main theorems (Theorems 1 and 2) are internally inconsistent with their proofs. Theorem 1's stated formula contradicts both the proof's final expression and the model's two-sided intersection count; Theorem 2's piecewise Z does not match the proof's geometry. These are not presentation typos but load-bearing errors: the applications in Section VI explicitly use Eq. (1). The paper would need a full rewrite of these derivations and a re-evaluation of the numerical results before it could be considered for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this paper aims at a real open problem—exact one-turn shortest-path distribution in an isotropic Poisson line Cox process—and the authors correctly note that prior work only covers Manhattan, means, or asymptotics. The claimed closed forms are new, and the proof strategy (void probability via geometric decomposition) is reasonable. But the central formulas as stated do not follow from the proofs, and the discrepancies are load-bearing.\n\nWhat's good: The problem is well-motivated; the comparison with the trivial recursive approximation in Sec. II is instructive; the applications to RIS and EV charging are plausible illustrations. The independence steps in the proofs are standard and mostly sound in spirit.\n\nThe soft spots are serious. Theorem 1's statement (Eq. 1) has exponent -2μt - 2λt + (λ/μ)(1-e^{-2μt}), but the proof immediately yields -2μt - 4λt + (2λ/μ)(1-e^{-2μt}). Remark 1 itself describes the term exp(-4λt) as the 'no line intersects' probability, consistent with the proof and not with Eq. (1). The proof also says N_t is Poisson with mean 2λt, but with the two-sided segment [-t,t] under the paper's Definition 1 the mean should be 4λt; either way, the stated formula is not what the proof gives. Theorem 2 has a similar mismatch: the piecewise Z in the statement (e.g., first case uses (sin ω - sin ω1)/sin(ω1-ω)) doesn't match the proof's Eq. (9), which uses (sin θ + sin θ1)/sin(θ1-θ). Theorem 3's bound is too under-specified to verify from the text. Since the applications in Section VI use Eq. (1) directly, the quantitative claims inherit the error.\n\nThis is not a case of a missing minor derivation; it's an internal contradiction between theorem statements and their own proofs. The paper could be salvageable—if the proof version is correct, the results would still be new—but as written the main exact results are unsupported. I would not cite the formulas in their current form. Still, the topic and approach merit a serious referee, because the errors are identifiable and potentially fixable, and the underlying problem is genuinely open.\n\nRecommendation: send to peer review only if the authors are given the opportunity to fix the discrepancies; otherwise, a desk reject with a clear explanation is defensible. My call: would accept for review, with the expectation of major revision.","headline":"The paper tackles a real open problem and the proof strategy is sound, but the main exact formulas contradict the proofs in a load-bearing way, so as written the results are unsupported.","tokens_in":12757,"tokens_out":2258,"would_cite":false,"duration_ms":19141,"reading_group":"no","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60D05","60G55"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper derives exact one-turn shortest-path distributions for Poisson line Cox processes, plus a computable two-turn bound, and applies them to V2V broadcasting and ride-hailing service guarantees.","keywords":["Poisson line process","Poisson line Cox process","shortest path length","void probability","stochastic geometry","vehicle-to-vehicle communication","ride-hailing","reconfigurable intelligent surface"],"falsifier":"Run a Monte Carlo simulation of a stationary Poisson line Cox process with fixed $\\lambda$ and $\\mu$, measure the empirical CCDF of the one-turn shortest path from a typical point, and compare it against Eq. (1) and against the same expression with $4\\lambda t$ in place of $2\\lambda t$; the matching variant identifies the correct intersection-count normalization.","tokens_in":11545,"feed_emoji":"🛣️","tokens_out":13591,"duration_ms":116151,"temperature":0.7,"pith_summary":"Closed-form distributions for street-constrained paths have been missing for the isotropic Poisson line Cox process, where random streets are drawn as a Poisson line process and users sit on each street as a one-dimensional Poisson process. This paper tries to close that gap. Under a one-turn restriction, it derives exact complementary CDFs for the nearest-neighbor path length from the typical point and from the typical intersection; under a two-turn restriction, it derives a computable upper bound that is tight in identified regimes. The formulas depend directly on the street density $\\lambda$ and the user density $\\mu$, and the paper uses them to evaluate the probability that a ride-hailing vehicle or a vehicle-to-vehicle link meets a service threshold, and to dimension electric-vehicle charging-point placement. A reader should care because these closed forms convert a quantity that was previously available only as a mean or through asymptotics into an explicit distribution that can be evaluated and compared across network parameters.","feed_headline":"Exact formulas for one-turn shortest paths on random street grids","feed_subtitle":"Closed-form trip-length distributions turn street and vehicle densities into service-guarantee probabilities.","key_machinery":"The load-bearing mechanism is the void-probability decomposition. The event $D>t$ is rewritten as the event that no point of the PLCP lies on the initial street within path length $t$, and no point lies on any crossing street within the remaining budget $t-s_i$ from its intersection with the initial street. Since the crossings of a Poisson line process with a fixed line form a one-dimensional Poisson process, the number of crossings is Poisson and their positions are uniform, so the union over all crossing lines becomes a product of exponentials that evaluates in closed form. For the typical intersection, the decomposition demands a case analysis of where a third line intersects the two streets through the typical intersection; the cases are summarized by the void segment length $Z(x,\\omega_1,\\omega)$ and averaged inside the integrals $T_x,T_y$. For the two-turn case, the proof shrinks the feasible set of line configurations and applies a conditional-expectation argument through $T(w,u)$, producing a nested exponential bound from the typical point.","core_discovery":"The paper's central claim is that the one-turn nearest-neighbor path length in a Poisson line Cox process has an exact, closed-form void probability. From the typical point, Theorem 1 states the CCDF $$\\bar F_D(t) = \\exp\\!\\left(-2\\mu t - 2\\$\\lambda$ t + \\frac{\\$\\lambda$}{\\mu}\\bigl(1-$e^{{-2\\mu t}}$\\bigr)\\right),$$ so that $F_D(t)=1-\\bar F_D(t)$. From the typical intersection, Theorem 2 gives a similar exact expression, $$F_D(t)=1-\\exp\\!\\left(-4\\mu t - 2\\$\\lambda$(2t-T_x-T_y)\\right),$$ where $T_x,T_y$ are triple integrals over the angles and crossing positions of a line that meets the two streets through the intersection, with integrand $\\exp(-\\mu Z(x,\\omega_1,\\omega))$ and $Z$ the length of the segment that must be empty of points. Theorem 3 bounds the two-turn CCDF from the typical point by a nested exponential whose innermost kernel $T(w,u)$ again enforces empty segments on the second street. If these expressions are correct, they give the first exact path-length distributions for an isotropic Poisson line Cox process, replacing earlier mean-value and asymptotic results with explicit functions of the line density $\\lambda$ and point density $\\mu$.","pith_inferences":["Extending the paper's approach, the same void-probability recursion should apply to any fixed number of turns as a deeper nesting of the same exponential factors; the general $k$-turn distribution is left open by the paper.","Fitted to a real street network, the two-parameter model provides a null benchmark: systematic deviations between measured trip-length distributions and these formulas would quantify anisotropy or non-Poissonian structure in the city grid.","The appearance of the ratio $\\lambda/\\mu$ in the exponents suggests that the shape of the distribution, not just its mean, changes across density regimes; the paper's own numerical comparison of the one-turn and two-turn curves makes these regime boundaries visible and testable.","In the V2V near-field setting, the additive RIS model turns the path-length CDF directly into a coverage probability, which implies a placement rule: favor intersection-mounted RIS when the one-turn intersection distribution dominates the two-turn point distribution."],"forward_implications":["With known street and user densities, the probability that the nearest reachable user lies within a target trip time is $F_D(t)$ from Theorem 1, giving a direct service-guarantee curve for ride-hailing.","In the RIS-assisted V2V application, the probability that the nearest vehicle decodes the safety message is $F_D$ evaluated at the SNR-derived path-length threshold, so link reliability becomes a closed-form function of $\\lambda$ and $\\mu$.","The parameter study shows regimes of dense users and sparse streets where the one-turn route from a typical intersection is stochastically shorter than the two-turn route from a typical point; this favors intersection-anchored placement of charging points and reflectors.","The two-turn upper bound, although an inequality, is tight in those regimes and provides a conservative reachability estimate for routing with at most two turns.","The lower and upper bounds for the typical-intersection case offer computationally cheap surrogates for system performance, with the lower bound tight for large $\\mu$ and the upper bound tight for large $\\lambda$."],"supporting_citations":[{"why":"supplies the Poisson count of intersections on a fixed line that Theorem 1's void probability is built on.","marker":"[4]"},{"why":"is the Manhattan-line version of the shortest-path problem whose methods the paper extends to isotropic orientations.","marker":"[10]"},{"why":"gives the earlier mean shortest-path computation via Neveu's exchange formula that the present distributional result refines.","marker":"[12]"},{"why":"is the exchange-formula tool that underlies the earlier mean path-length approach.","marker":"[13]"},{"why":"provides asymptotic scaling limits for shortest paths in stationary tessellations, the large-network limit the paper makes exact at finite t.","marker":"[14]"},{"why":"supplies the near-field RIS path-loss model used in the V2V communication application.","marker":"[18]"},{"why":"motivates the electric-vehicle charging-point placement application addressed with the derived distributions.","marker":"[20]"}],"fun_headline_variants":["Exact one-turn path length formulas for Poisson line Cox","Closed-form trip lengths: street density to service guarantees","One-turn shortest paths on random grids: exact distributions","Poisson street networks: exact path length CDFs","From street density to trip-time guarantees: exact formulas"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof of Theorem 1 assumes that the number of street crossings on the initial street within distance $t$ of the typical point is Poisson with mean $2\\lambda t$; if both sides of the point should be counted, the correct mean is $4\\lambda t$, and the exponents in the main formulas must be rescaled.","fun_headline_variants_meta":{"raw":{"variants":["Exact one-turn path length formulas for Poisson line Cox","Closed-form trip lengths: street density to service guarantees","One-turn shortest paths on random grids: exact distributions","Poisson street networks: exact path length CDFs","From street density to trip-time guarantees: exact formulas"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000182,"raw_usage":{"total_tokens":1361,"prompt_tokens":1043,"completion_tokens":318,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":659,"completion_tokens_details":{"reasoning_tokens":240}},"tokens_in":659,"tokens_out":318,"duration_ms":3363,"temperature":1.0,"reasoning_tokens":240,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T13:09:55.588436+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a Monte Carlo simulation of a stationary Poisson line Cox process with fixed $\\lambda$ and $\\mu$, measure the empirical CCDF of the one-turn shortest path from a typical point, and compare it against Eq. (1) and against the same expression with $4\\lambda t$ in place of $2\\lambda t$; the matching variant identifies the correct intersection-count normalization.","supporting_citations":[{"cited_title":"Poisson line Cox process: Foundations and applications to vehicular networks,","cited_arxiv_id":null,"evidence_quote":"supplies the Poisson count of intersections on a fixed line that Theorem 1's void probability is built on."},{"cited_title":"Shortest path distance in Manhattan Poisson line Cox process,","cited_arxiv_id":null,"evidence_quote":"is the Manhattan-line version of the shortest-path problem whose methods the paper extends to isotropic orientations."},{"cited_title":"Analysis of shortest paths and subscriber line lengths in telecommunication access networks,","cited_arxiv_id":null,"evidence_quote":"gives the earlier mean shortest-path computation via Neveu's exchange formula that the present distributional result refines."},{"cited_title":"Neveu’s exchange formula for analysis of wireless net- works with hotspot clusters,","cited_arxiv_id":null,"evidence_quote":"is the exchange-formula tool that underlies the earlier mean path-length approach."},{"cited_title":"Scaling limits for shortest path lengths along the edges of stationary tessellations,","cited_arxiv_id":null,"evidence_quote":"provides asymptotic scaling limits for shortest paths in stationary tessellations, the large-network limit the paper makes exact at finite t."},{"cited_title":"Wireless communications with reconfigurable intelligent surface: Path loss modeling and experimental measurement,","cited_arxiv_id":null,"evidence_quote":"supplies the near-field RIS path-loss model used in the V2V communication application."},{"cited_title":"Stochastic geometry planning of electric vehicles charging stations,","cited_arxiv_id":null,"evidence_quote":"motivates the electric-vehicle charging-point placement application addressed with the derived distributions."}],"review_version":1}