Signal Processing · arXiv:2611.16288 v1 · submitted 12 Nov 2026 · 15 references
Cost of Decibel-Domain Averaging of Line-of-Sight States in Aerial Coverage Analysis
The paper says
“Uncrewed aerial vehicle (UAV) coverage studies routinely average the line-of-sight (LoS) and non-line-of-sight (NLoS) path loss in the decibel domain, weighted by their probabilities. Since the conditional path-loss distribution is bimodal, the resulting mean path loss describes no physical link.”
more of the abstract
We give the error of this mean path loss, relative to the path loss that corresponds to the mean received power, in closed form. We quantify it by ray tracing an ITU-R P.1410 city, supplying the averaged model with the exact LoS probability so that no parameter error contributes. Our numerical results show that the averaged path loss is 6.6 dB higher than the path loss of the mean received power and lies within 2 dB of only 6.2% of the links. It moves the coverage-optimal altitude from 1764 m to 778 m and understates the outage probability by up to 0.198.
from the abstract
Pith says
Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded · not checked by a person
It can be wrong. Challenge this review or reply to it.
The passes disagreed. Reader: holds up. Skeptic: holds up only with conditions. Referee: major revision.
Correct, honest, and useful: a closed-form proof that decibel-domain averaging of LoS/NLoS path loss biases downstream coverage quantities, with the biggest claimed numbers hostage to one disclosed modeling choice. This line is wrong
3 major objections · 8 minor
How often objections from reviews like this one are right: not yet measured
Not checked: figures, code, what the 15 references say.
The paper says, at Section III, line 70
The review did not quote this passage.
Pith says: Objection 1 of 3 · major
The paper supplies the decibel-averaged model with a shadowing spread defined as sigma_bar = PLoS * sigma_LoS + (1-PLoS) * sigma_NLoS (line 70). This interpolation is not part of the original method in the cited works (Table I). Those works typically use the averaged path loss to define a coverage radius or SNR threshold without an explicit shadowing spread, or they apply a single fixed sigma.
more of objection 1
By inventing this specific sigma interpolation, the paper may be comparing against a version of the averaged model that is more vulnerable to error than any version actually used in practice. The claims about outage error (up to 0.198) and optimal altitude shift (factor 2.3) depend on this choice. The paper should justify this interpolation or show that the results are robust to alternative choices (e.g., using sigma_NLoS for all links, or using a single sigma from the literature).
The paper says, at Section IV, Table II
The review did not quote this passage.
Pith says: Objection 2 of 3 · major
The paper reports that the mixture model (6) reproduces the ray-traced outage to within 0.004 and finds the same optimum altitude in all four environments. However, the mixture model uses the same excess losses and shadowing standard deviations as the ray trace.
more of objection 2
This is a consistency check, not a validation. The ray trace itself uses these parameters (eta_LoS=1 dB, eta_NLoS=20 dB, sigma_LoS=4 dB, sigma_NLoS=8 dB) from Al-Hourani et al. (2014). The mixture model is essentially a parametric fit to the same data. The fact that it matches the ray trace is expected, not a demonstration of superiority. The paper should clarify that the comparison is between two models that share the same inputs, and that the ray trace is the reference, not an independent validation of the mixture model.
The paper says, at Section IV, Fig. 2(c) and Table II
The review did not quote this passage.
A classifier read this sentence of the paper as acknowledging the concern (probability 0.70):
“An outage probability from (2) also needs a shadowing spread, which (2) does not define because it represents both states by a single mean.”
Pith says: Objection 3 of 3 · major
The paper states that the decibel-averaged model understates outage probability by up to 0.198 and shifts the optimal altitude from 1764 m to 778 m. However, the decibel-averaged model is not designed for outage calculations; it is typically used to define a coverage radius or an SNR threshold (as noted in Table I).
more of objection 3
The paper's construction of an outage probability for the averaged model (using sigma_bar) is not standard. The paper should discuss whether the cited works actually compute outage probabilities from the averaged path loss, or whether they use it for other purposes (e.g., coverage radius). If the latter, the relevance of the outage comparison is limited.
8 minor objections
The paper says, at Abstract
Pith says: minor objection 1 of 8
The abstract states that the averaged path loss is '6.6 dB higher than the path loss of the mean received power.' This is for a specific geometry (500 m range, 300 m altitude) and parameter set. The paper should note that this value is representative, not universal.
The paper says, at Section II, Eq. (4)
Pith says: minor objection 2 of 8
The derivation of Eq. (4) is clear, but it would be helpful to explicitly state that PLoS cancels the free-space path loss, so the error depends only on the excess losses and PLoS.
show 6 more
The paper says, at Section II, Eq. (5)
Pith says: minor objection 3 of 8
The asymptotic approximations are useful. The paper should note the range of validity more explicitly (e.g., 'for Delta_eta < 5 dB' for the small-Delta approximation).
The paper says, at Section III, Fig. 2(a)
Pith says: minor objection 4 of 8
The paper compares the published sigmoid (a=9.61, b=0.16) with the ray-traced PLoS and finds a large discrepancy (up to 0.33). This is interesting but somewhat tangential to the main claim about averaging error. The paper should clarify that this discrepancy is due to the specific city parameters and user placement, not a general flaw in the sigmoid model.
The paper says, at Section IV, Fig. 2(b)
Pith says: minor objection 5 of 8
The paper states that only 6.2% of links have path loss within 2 dB of the averaged value. This is a striking statistic, but it depends on the bimodality of the distribution. The paper should note that this percentage would be higher if the modes were closer together (e.g., suburban environment).
The paper says, at Section IV, Table II
Pith says: minor objection 6 of 8
The table shows that the decibel average selects an altitude h_dB that is lower than h* by a factor between 1.8 and 5.7. The paper should explain why the factor is largest (5.7) in the high-rise urban environment, where PLoS is lowest (0.08).
The paper says, at Section V, Conclusion
Pith says: minor objection 7 of 8
The conclusion recommends three changes: use (3) for mean received power, (6) for outage, and (7) for rate. This is sound advice, but the paper should acknowledge that these changes require knowledge of the state-dependent shadowing standard deviations, which may not always be available.
The paper says, at General
Pith says: minor objection 8 of 8
The paper is well-written and the main argument is clear. The figures are informative. The references are appropriate.
What would settle it
Recompute the outage curves in Fig. 2(c) and Table II for the decibel-averaged model using two alternative shadowing treatments: (i) a single fixed sigma = sigma_NLoS = 8 dB (the more conservative choice that a cautious designer might adopt), and (ii) the mixture model's own sigma values but applied to the averaged mean (i.e., compute outage as PLoS * Q((PL_max - PL_bar)/sigma_LoS) + (1-PLoS) * Q((PL_max - PL_bar)/sigma_NLoS), which is a hybrid that keeps the averaged mean but respects the separate spreads). If the outage error and altitude shift shrink substantially under either alternative, then the paper's headline numerical results are artifacts of the specific sigma interpolation chosen, not of the decibel-domain averaging per se.
from the skeptic's reading
How the review went
4 readings by deepseek-v4-flash · 13 Nov 2026
- Reader holds up
- Skeptic partly agrees: says it holds up only with conditions Ray-traced PLoS is supplied to the averaged model, but the averaged model's outage calculation still requires a shadowing spread; the paper's choice of a probability-weighted linear interpolation of sigma (line 70) is ad hoc and may systematically favor the mixture model.
- Editor wrote the verdict above
- Referee major revision: 3 major objections, 8 minor
Not run on this paper: the simulated authors' defence.
Pith did not check this review's objections against the paper.
The review in full
The review in brief, machine-written
Short version: this is a correct, honest, useful letter. It proves what people in UAV coverage analysis have suspected — that decibel-domain averaging of LoS/NLoS path loss (Eq. 2) describes no physical link — and turns the suspicion into a closed-form error term. I re-derived Eq. 4 by hand at the representative link (PLoS = 0.495, eta_LoS = 1, eta_NLoS = 20): it returns 6.6 dB, matching the paper. The math is elementary Jensen-gap territory, but as a tool it's new and it's right.
The design is the best part. By feeding the averaged model the exact ray-traced PLoS rather than a fitted sigmoid, the paper isolates the averaging error from parameter error. That control is clean. The four-environment table (suburban through high-rise) shows the effect isn't an artifact of one city, and the rate calculation (Fig. 1c, up to 12.5% low) adds a useful result with the opposite sign. Disclosure is transparent, including the AI-assistance note, and the author's own earlier survey is cited honestly as the qualitative precursor.
Soft spots, in order of size.
The stress-test concern is partly fair. Eq. (2) defines no shadowing spread, so the paper supplies an interpolated sigma (line 70), and the authors say so explicitly. That means the headline outage number (0.198) and the altitude shift (778 m vs 1764 m) ride on a modeling choice a practitioner isn't forced to make. But the stress-test overstates the fragility: the altitude error is mostly driven by the mean path loss crossing the link budget, a property of (2) itself, so "optimal altitude is too low" likely survives a different sigma. A sensitivity check — sigma fixed at sigma_NLoS, or coverage judged by the mean against budget — would settle it.
The other limit is the acknowledged one: excess-loss and shadowing values are representative, not site-specific (b130). The paper says this and shows how the error scales with mode separation, so it doesn't overclaim.
Who this is for: anyone doing optimal-altitude, coverage-radius, or outage work with the Al-Hourani model line. It's a methodology warning with concrete corrections (Eqs. 3, 6, 7).
Recommendation: send it to review. Correct, short, and it hands the community a closed-form error term instead of hand-waving. The referee should ask for the sigma sensitivity check; that's a minor revision, not a rejection.
The strongest objection
The central controlled experiment (Section III, Fig. 2(c)) compares the mixture model (6) against the decibel-averaged model (2) under the same exact PLoS. However, the decibel-averaged model does not naturally define a shadowing spread because it collapses the two states into one mean. The paper supplies it with a probability-weighted linear interpolation of the two sigma values: sigma_bar = PLoS * sigma_LoS + (1-PLoS) * sigma_NLoS (line 70). This choice is not derived from any physical or statistical principle; it is simply the same linear weighting used for the means. The outage probability of the averaged model is then computed as Q((PL_max - PL_bar)/sigma_bar). This construction is not the standard use of the decibel-averaged model in the cited literature (Table I). Those works typically use the averaged path loss to define a coverage radius or an SNR threshold without any explicit shadowing spread, or they apply a single fixed shadowing sigma. By inventing a specific sigma interpolation that is not part of the original method, the paper may be comparing against a straw-man version of the averaged model that is more vulnerable to the error than any version actually used in practice. The claim that the averaged model understates outage by up to 0.198 and shifts the optimal altitude by a factor of 2.3 depends on this particular sigma interpolation. If a practitioner using the averaged model instead applied a single conservative sigma (e.g., sigma_NLoS = 8 dB for all links), the outage error and altitude shift could be smaller. The paper does not test this sensitivity.
The reader's weakest_assumption focuses on the representativeness of the ITU-R P.1410 city parameters and the specific excess loss and shadowing values. That is a valid concern, but the paper acknowledges it (b130) and provides a closed-form error expression (Eq. 4) that is parameter-independent. My concern is more fundamental: the controlled experiment that isolates the averaging error from parameter error introduces an additional, unexamined modeling choice for the shadowing spread of the averaged model. The paper states (line 70): 'We therefore model the path loss of the averaged model as one Gaussian with mean PL and standard deviation sigma_bar = PLoS * sigma_LoS + (1-PLoS) * sigma_NLoS, which interpolates the two spreads in the same way as (2) interpolates the two means.' This is presented as a natural extension, but it is not part of the original method (2) as used in the cited works. The original method (2) produces a single path loss value; it does not specify how to compute outage from that value when shadowing is present. The cited works either ignore shadowing entirely (using a deterministic path loss threshold) or apply a single shadowing margin. By constructing a specific sigma interpolation that is linear in probability, the paper may be creating a version of the averaged model that is particularly bad at capturing the NLoS tail, because the interpolated sigma is always less than sigma_NLoS when PLoS > 0. This systematically reduces the outage probability computed from the averaged model, making the mixture model look better by comparison. The paper does not test whether this interpolation is the fairest comparison. The concrete test I propose would determine whether the numerical results are robust to the sigma interpolation choice. If they are not, then the central quantitative claims (6.6 dB, 1764 m vs 778 m, outage error 0.198) are not solely attributable to the decibel-domain averaging but depend on an auxiliary modeling decision. The paper's qualitative point that decibel-domain averaging is problematic would still stand, but the specific magnitudes would be called into question. I recommend CONDITIONAL acceptance: the paper should either justify the sigma interpolation as the standard practice in the cited literature (which it does not) or test the sensitivity of its results to this choice. The paper is otherwise well-argued and the closed-form error expression (Eq. 4) is a valuable contribution independent of the numerical results.
The referee's report
If the results hold, the paper identifies a systematic error in a widely used simplification in UAV coverage analysis. The closed-form error expression (Eq. 4) is a useful tool for practitioners to assess the cost of the approximation. The ray-traced evaluation with controlled PLoS isolates the averaging error from parameter errors, which is a clean experimental design. The results are likely to influence how future work models path loss in aerial networks.
Pith's reading of the paper (a machine's summary, not the authors')
This paper argues that a widely used averaging method in UAV coverage analysis—weighting line-of-sight (LoS) and non-line-of-sight (NLoS) path losses by their probabilities in the decibel domain—produces a mean path loss that does not correspond to any physical link. Because the conditional path-loss distribution is bimodal, the averaged value falls in the valley between the two modes, where no actual link lies. The author derives a closed-form expression for the resulting error and quantifies it using ray-traced simulations of an ITU-R P.1410 city. The numerical results show that the averaged path loss is 6.6 dB higher than the path loss of the mean received power, lies within 2 dB of only 6.2% of links, shifts the coverage-optimal altitude from 1764 m to 778 m, and understates outage probability by up to 0.198. The paper proposes simple corrections: averaging in the linear domain for mean received power, and mixing conditional tail probabilities for outage and rate calculations.
What carries the argument. The central object is the averaging error function Δ (Equation 4), which gives the difference between the decibel-domain weighted mean path loss and the path loss of the mean received power. It depends only on the LoS probability p and the two excess losses η_LoS and η_NLoS, and has closed-form limiting behavior: quadratic in the mode separation for small Δ_η, linear for large Δ_η. The paper uses this function to predict the error magnitude without simulation, and then validates it against a ray-traced ITU-R P.1410 city where the exact LoS probability is known from geometry.
Core claim. The paper establishes that decibel-domain averaging of LoS and NLoS path losses (Equation 2) introduces a systematic error because the logarithm is concave and the two-state link distribution is bimodal. The error has a closed form (Equation 4) that depends only on the LoS probability and the two excess losses. In a ray-traced ITU-R P.1410 urban city with standard parameters, the averaged path loss is 6.6 dB higher than the path loss corresponding to the mean received power (Equation 3). Supplying the averaged model with the exact LoS probability from ray tracing isolates this averaging error from any parameter error in the LoS probability model itself. The consequence is that coverage-optimal altitudes are systematically underestimated (by factors of 1.8 to 5.7 across four environments), and outage probabilities are understated by up to 0.23. The paper demonstrates that a two-state mixture model (Equation 6 for outage, Equation 7 for rate) removes the error and reproduces ray-traced results to within 0.004 in outage probability.
The premise it rests on. The numerical magnitudes reported (6.6 dB, 1764 m vs 778 m, etc.) depend on the specific excess losses (η_LoS=1 dB, η_NLoS=20 dB) and shadowing standard deviations (σ_LoS=4 dB, σ_NLoS=8 dB) from Al-Hourani et al. (2014), which the paper acknowledges are representative rather than site-specific; the closed-form error expression is general, but the quantitative conclusions about optimal altitude shifts and outage understatement are tied to these particular values.
If the paper is right (Pith's inferences, not the authors')
Any aerial coverage study that uses Equation 2 to compute coverage radius, optimal altitude, or outage probability will systematically underestimate the optimal altitude and understate outage probability, regardless of improvements to the LoS probability model.
The closed-form error expression (Equation 4) allows researchers to correct existing results retroactively if they have the LoS probability and excess losses used in their study.
The mixture model (Equation 6) for outage and (Equation 7) for rate can be adopted as a drop-in replacement that requires no additional parameters beyond those already used in the averaged model.
The paper's controlled experiment—supplying the averaged model with the exact ray-traced LoS probability—shows that the error is intrinsic to the averaging step and cannot be fixed by better LoS probability models alone.
The same decibel-domain averaging error likely affects other bimodal propagation models, such as those for indoor/outdoor transitions or terrestrial links with foliage blockage, where the conditional path-loss distribution is also bimodal.
The paper's results imply that existing UAV deployment algorithms trained on coverage maps generated with Equation 2 may be optimizing for the wrong altitude, potentially explaining discrepancies between simulation and field trials.
A measured dataset with paired link-state and path-loss recordings, as the paper calls for, could validate the closed-form error expression in real environments and establish whether the representative excess losses used here are conservative or optimistic.
The error's dependence on mode separation suggests that in millimeter-wave bands where η_NLoS is much larger than η_LoS, the averaging error could exceed 17 dB, making the decibel-domain average even less representative.
Check this review
Pith stored no record of this review's passes and took no fingerprint of it. Reviews made by code from 13 Nov 2026 on carry both.
How often reviews like this one are right
Reviews like this one: written by deepseek-v4-flash, where what was read was not recorded.
How often objections from reviews like this one are right: not yet measured
Reviews by people
This review was written by a machine. No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.
The paper
Abstract
Uncrewed aerial vehicle (UAV) coverage studies routinely average the line-of-sight (LoS) and non-line-of-sight (NLoS) path loss in the decibel domain, weighted by their probabilities. Since the conditional path-loss distribution is bimodal, the resulting mean path loss describes no physical link. We give the error of this mean path loss, relative to the path loss that corresponds to the mean received power, in closed form. We quantify it by ray tracing an ITU-R P.1410 city, supplying the averaged model with the exact LoS probability so that no parameter error contributes. Our numerical results show that the averaged path loss is 6.6 dB higher than the path loss of the mean received power and lies within 2 dB of only 6.2% of the links. It moves the coverage-optimal altitude from 1764 m to 778 m and understates the outage probability by up to 0.198.
The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. No sentence in this record was written by a model.
abstract compiled · 2026-11-13
The authors' main claim, as the classifier picked it out 0.43
“Since the conditional path-loss distribution is bimodal, the resulting mean path loss describes no physical link.”
How the main claim is stated: asserted 1.00
What the abstract itself states, as probabilities: claims first 0.05 · states limitation 0.20 · human participants 0.01 · controlled comparison 0.42 · randomized assignment 0.01 · states quantitative result 0.99 · compares with prior results 0.71 · mentions code or data availability 0.03
The abstract, sentence by sentence
This table labels the abstract only (arXiv listing abstract v1). The number is the probability the classifier gave the label.
| Sentence | Role | The authors' own work | States the main conclusion | Strength |
|---|---|---|---|---|
| Uncrewed aerial vehicle (UAV) coverage studies routinely average the line-of-sight (LoS) and non-line-of-sight (NLoS) path loss in the decibel domain, weighted by their probabilities. | background 1.00 | 0.06 | 0.03 | asserted 1.00 |
| Since the conditional path-loss distribution is bimodal, the resulting mean path loss describes no physical link. | background 0.56 | 0.14 | 0.31 | asserted 1.00 |
| We give the error of this mean path loss, relative to the path loss that corresponds to the mean received power, in closed form. | contribution 0.99 | 0.95 | 0.14 | asserted 1.00 |
| We quantify it by ray tracing an ITU-R P.1410 city, supplying the averaged model with the exact LoS probability so that no parameter error contributes. | method 0.99 | 0.95 | 0.05 | asserted 0.99 |
| Our numerical results show that the averaged path loss is 6.6 dB higher than the path loss of the mean received power and lies within 2 dB of only 6.2% of the links. | result 1.00 | 0.96 | 0.47 | asserted 1.00 |
| It moves the coverage-optimal altitude from 1764 m to 778 m and understates the outage probability by up to 0.198. | result 0.90 | 0.93 | 0.37 | asserted 1.00 |
What kind of paper the classifier reports this is: theoretical result 0.93 · evidence from simulation or computation 0.98
What was examined, and what was not
Sources read
- arXiv listing abstract v1
Not examined
- full text
- figures and tables
- references and citing sentences
- relations to other papers
Levels run
- P abstract
Not run
- S sections
- T body sentences
- C citations
- R cross-paper relations
How the labels were made
Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-11-13: 1 request, 5,010 input tokens.
Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. This record as JSON
Figures (2)
References (15)
- [1] Y . Zeng, Q. Wu, and R. Zhang, “Accessing from the sky: A tutorial on UA V communications for 5G and beyond,”Proc. IEEE, vol. 107, no. 12, pp. 2327–2375, Dec. 2019 on Pith
- [2] A. Al-Hourani, S. Kandeepan, and S. Lardner, “Optimal LAP altitude for maximum coverage,”IEEE Wireless Commun. Lett., vol. 3, no. 6, pp. 569–572, Dec. 2014 on Pith
- [3] ITU-R, “Propagation data and prediction methods required for the design of terrestrial broadband radio access systems operating in a frequency range from 3 to 60 GHz,” International Telecommunication Union, Tech. Rep. Recommendation P.1410-6, Aug. 2023 on Pith
- [4] R. I. Bor-Yaliniz, A. El-Keyi, and H. Yanikomeroglu, “Efficient 3-D placement of an aerial base station in next generation cellular networks,” inProc. IEEE Int. Conf. Commun. (ICC), Kuala Lumpur, Malaysia, May 2016, pp. 1–5 on Pith
- [5] M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, “Drone small cells in the clouds: Design, deployment and performance analysis,” inProc. IEEE Global Commun. Conf. (GLOBECOM), San Diego, CA, Dec. 2015, pp. 1–6 on Pith
- [6] M. Alzenad, A. El-Keyi, F. Lagum, and H. Yanikomeroglu, “3-D place- ment of an unmanned aerial vehicle base station (UA V-BS) for energy- efficient maximal coverage,”IEEE Wireless Commun. Lett., vol. 6, no. 4, pp. 434–437, Aug. 2017 on Pith
- [7] M. Alzenad, A. El-Keyi, and H. Yanikomeroglu, “3-D placement of an unmanned aerial vehicle base station for maximum coverage of users with different QoS requirements,”IEEE Wireless Commun. Lett., vol. 7, no. 1, pp. 38–41, Feb. 2018 on Pith
- [8] Y . Wang and J. Farooq, “Deep-reinforcement-learning-based placement for integrated access backhauling in UA V-assisted wireless networks,” IEEE Internet Things J., vol. 11, no. 8, pp. 14 727–14 738, Apr. 2024 on Pith
- [9] A. Vavoulas, K. K. Delibasis, H. G. Sandalidis, G. Nousias, and N. Vaiopoulos, “Efficient UA V coverage in large convex quadrilateral areas with elliptical footprints,”IEEE Wireless Commun. Lett., vol. 15, pp. 1479–1483, 2026 on Pith
- [10] B. A. Homssi and A. Al-Hourani, “Optimal beamwidth and altitude for maximal uplink coverage in satellite networks,”IEEE Wireless Commun. Lett., vol. 11, no. 4, pp. 771–775, Apr. 2022 on Pith
- [11] W. Khawaja, I. Guvenc, D. W. Matolak, U.-C. Fiebig, and N. Schneck- enburger, “A survey of air-to-ground propagation channel modeling for unmanned aerial vehicles,”IEEE Commun. Surveys Tuts., vol. 21, no. 3, pp. 2361–2391, 3rd Quart. 2019 on Pith
- [12] A. Al-Hourani, “On the probability of line-of-sight in urban environ- ments,”IEEE Wireless Commun. Lett., vol. 9, no. 8, pp. 1178–1181, Aug. 2020 on Pith
- [13] A. Saboor, E. Vinogradov, Z. Cui, A. Al-Hourani, and S. Pollin, “A geometry-based modelling approach for the line-of-sight probability in UA V communications,”IEEE Open J. Commun. Soc., vol. 5, pp. 364– 378, 2024 on Pith
- [14] L. Chen, W. Zhang, M. A. Kishk, and M.-S. Alouini, “Correlation of line-of-sight probabilities in aerial-terrestrial communications: Model- ing, analysis, and application,”IEEE Trans. V eh. Technol., vol. 73, no. 5, pp. 7049–7065, May 2024 on Pith
- [15] J. A. Rice,Mathematical Statistics and Data Analysis, 3rd ed. Belmont, CA: Brooks/Cole, 2007, sec. 4.6 on Pith
Challenges and discussion
Disagree with the review, or know something it missed? Leave a comment, or challenge this review formally: a challenge becomes a permanent entry in this paper's record, and its decision is published here.
Come back by email
A short list of the week's newly reviewed papers. Nothing is sent until you confirm from your inbox, and every email has a link that stops it. No account.


No comments yet.
Sign in to comment