pith. sign in

arxiv: 2510.04346 · v2 · submitted 2025-10-05 · 💻 cs.NI · cs.LG· cs.NA· eess.SP· math.NA

Environment-Aware Indoor LoRaWAN Path Loss: Parametric Regression Comparisons, Shadow Fading, and Calibrated Fade Margins

Nahshon Mokua Obiri , Kristof Van Laerhoven This is my paper

Pith reviewed 2026-05-18 10:15 UTC · model grok-4.3

classification 💻 cs.NI cs.LGcs.NAeess.SPmath.NA
keywords LoRaWANpath lossindoor propagationenvironmental covariatesfade marginshadow fadingparametric regressionIoT reliability
0
0 comments X

The pith

Incorporating indoor environmental readings into LoRaWAN path loss models improves prediction accuracy and lowers the fade margin needed for 99 percent reliability.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a path loss model for indoor LoRaWAN that adds time-varying environmental measurements such as humidity, temperature, CO2, particulates, and pressure to a standard log-distance multi-wall baseline. On a twelve-month dataset from one office, a selective polynomial regression on the continuous environmental variables outperforms linear alternatives in time-blocked cross-validation. The gain in accuracy lets the model prescribe smaller fade margins that still meet a one-percent outage target on held-out data, which matters for planning energy-efficient IoT networks. The study further shows that the residuals are not well described by a single normal distribution but instead by a compact three-component Gaussian mixture.

Core claim

An environment-conditioned path loss framework augments a log-distance multi-wall baseline with co-recorded covariates (relative humidity, temperature, carbon dioxide, particulate matter, barometric pressure) and SNR; a selective quadratic extension on the continuous predictors reduces cross-validated RMSE from 8.23 dB to 7.38 dB and raises R-squared from 0.81 to 0.84, while out-of-fold residuals are best summarized by a three-component Gaussian mixture whose upper-tail percentiles supply calibrated fade margins that achieve the target outage rate at lower values than linear baselines.

What carries the argument

environment-conditioned path loss framework that augments a log-distance multi-wall baseline with environmental covariates and applies selective polynomial regression

If this is right

  • The selective polynomial model reduces cross-validated root mean square error to 7.38 dB from 8.23 dB and increases R-squared to 0.84 from 0.81.
  • Out-of-fold residuals follow a three-component Gaussian mixture with a sharp core and light broad tail rather than a single normal distribution.
  • At a one-percent outage target the polynomial model requires a 25.73 dB fade margin versus 27.79 to 28.05 dB for linear baselines.
  • The resulting link budgets become tighter and more aligned with sixth-generation reliability targets under energy constraints for massive indoor IoT.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same environmental-conditioning approach could be tested on other indoor wireless technologies to see whether similar accuracy gains appear.
  • Deployments in changing indoor spaces may need periodic re-collection of paired signal and environment data to keep the model current.
  • Real-time fusion of environmental sensor streams with signal reports could support dynamic adjustment of transmit power or margins.

Load-bearing premise

The relationships between the recorded environmental covariates and path loss observed in one eighth-floor office over twelve months will hold in other indoor spaces and time periods without retraining.

What would settle it

New measurements collected in a different indoor environment where the selective polynomial model fails to reduce RMSE below the linear baselines or where the prescribed fade margins no longer achieve the one-percent outage target on the new data.

read the original abstract

Indoor long range wide area network (LoRaWAN) propagation is shaped by structural and time-varying environmental factors, which limit single-slope log-distance models and the standard log-normal shadowing assumption. We propose an environment-conditioned path loss framework that augments a log-distance multi-wall baseline with co-recorded environmental covariates (relative humidity, temperature, carbon dioxide, particulate matter, and barometric pressure) and receiver-reported signal-to-noise, and we validate both the mean and the residual law statistically. The approach is evaluated on a 12-month campaign in an eighth-floor office (240 m^2) using time-blocked 5-fold cross-validation and a chronological hold-out. Across parametric regressors (regularized multiple linear regression (MLR), conjugate Bayesian linear regression, and a selective quadratic MLR extension on continuous predictors), the selective polynomial mean improves out-of-sample accuracy, reducing cross-validated root mean square error from 8.23 to 7.38 dB and increasing R^2 from 0.81 to 0.84. Out-of-fold (OOF) residuals are distinctly non-Gaussian and are best summarized by a compact 3-component Gaussian mixture with a sharp core and a light, broad tail. Finally, we translate prediction error into reliability by prescribing the fade margin as the upper-tail percentile of OOF errors, attaching moving-block bootstrap uncertainty, and validating the resulting outage on a held-out set. At a 1% outage target (99% reliability), the polynomial model requires 25.73 dB versus 27.79 to 28.05 dB for linear baselines, enabling tighter indoor massive Internet of Things link budgets aligned with sixth-generation reliability targets under energy constraints.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes an environment-conditioned path loss framework for indoor LoRaWAN that augments a log-distance multi-wall baseline with co-recorded environmental covariates (relative humidity, temperature, carbon dioxide, particulate matter, barometric pressure) and receiver SNR. It compares parametric regressors (regularized MLR, conjugate Bayesian linear regression, and a selective quadratic MLR extension on continuous predictors) on a 12-month campaign in a single 240 m² eighth-floor office using time-blocked 5-fold cross-validation and a chronological hold-out. The selective polynomial model reduces cross-validated RMSE from 8.23 dB to 7.38 dB and raises R² from 0.81 to 0.84. Out-of-fold residuals are modeled with a 3-component Gaussian mixture, and fade margins are prescribed as upper-tail percentiles of OOF errors with moving-block bootstrap uncertainty, validated on the held-out set to achieve 25.73 dB at 1% outage versus 27.79–28.05 dB for linear baselines.

Significance. If the reported gains in mean prediction accuracy and tail reliability hold under the environmental statistics of the studied office, the framework could support tighter indoor LoRaWAN link budgets for massive IoT deployments aligned with sixth-generation reliability targets. Strengths include the use of time-blocked CV plus chronological hold-out to mitigate overfitting, bootstrap uncertainty on fade margins, and explicit statistical modeling of non-Gaussian residuals. The single-site data collection, however, limits immediate broader impact until transferability is addressed.

major comments (2)
  1. [Evaluation section] Evaluation section: The headline improvements (RMSE reduction 8.23→7.38 dB, R² increase 0.81→0.84, and 1% outage fade margin of 25.73 dB) are obtained exclusively from time-blocked CV and hold-out within one 240 m² eighth-floor office campaign. Because the selective quadratic extension conditions on site-specific covariates whose marginal effects may be entangled with fixed structural factors (layout, HVAC schedule, sensor placement), the reported advantage does not automatically extend to other indoor spaces without re-fitting or additional validation data.
  2. [Fade margin calibration] Fade margin calibration: The 1% outage margins are defined as upper-tail percentiles of the out-of-fold residuals from the very models under comparison. Although a chronological hold-out is used for final outage validation, this construction still makes the margin a direct function of the fitted prediction errors, which could affect the claimed independence of the reliability assessment.
minor comments (2)
  1. [Abstract] The abstract states that 'statistical checks on residuals and outage are mentioned' but does not name the exact tests or data exclusion rules; adding one sentence with the specific procedures would improve clarity.
  2. [Methods] Notation for the selective quadratic extension (which covariates receive quadratic terms and how selection is performed) should be defined explicitly in the methods section rather than left to the results tables.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed feedback on our manuscript. We address each major comment below and describe the corresponding revisions.

read point-by-point responses

  1. Referee: [Evaluation section] Evaluation section: The headline improvements (RMSE reduction 8.23→7.38 dB, R² increase 0.81→0.84, and 1% outage fade margin of 25.73 dB) are obtained exclusively from time-blocked CV and hold-out within one 240 m² eighth-floor office campaign. Because the selective quadratic extension conditions on site-specific covariates whose marginal effects may be entangled with fixed structural factors (layout, HVAC schedule, sensor placement), the reported advantage does not automatically extend to other indoor spaces without re-fitting or additional validation data.

    Authors: We agree that all quantitative results are obtained from a single 240 m² office environment and that the reported gains may not transfer directly to other indoor spaces without re-calibration. The environmental covariates can interact with fixed building features, so the advantage is site-conditioned by construction. In the revised manuscript we will add a new subsection in the Discussion that explicitly states this limitation, notes the need for site-specific re-fitting, and outlines possible transfer-learning directions for future multi-site studies. revision: yes

  2. Referee: [Fade margin calibration] Fade margin calibration: The 1% outage margins are defined as upper-tail percentiles of the out-of-fold residuals from the very models under comparison. Although a chronological hold-out is used for final outage validation, this construction still makes the margin a direct function of the fitted prediction errors, which could affect the claimed independence of the reliability assessment.

    Authors: We acknowledge that the fade margins are derived from the empirical distribution of out-of-fold residuals and are therefore model-dependent by design. The chronological hold-out set, however, remains completely unseen during model fitting and cross-validation and is used solely to verify that the prescribed margins produce the target outage rates. In the revision we will clarify this distinction in the text, stating that while the numerical margin values depend on the fitted residuals, their reliability is externally validated on independent data. We view the current procedure as standard for empirical link-budget design but will improve the wording to avoid any implication of full statistical independence. revision: partial

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper's methodology relies on standard empirical validation: time-blocked 5-fold cross-validation combined with a chronological hold-out set to assess mean prediction accuracy (RMSE, R²) and to derive fade margins from out-of-fold residuals, with explicit outage validation on the held-out data. This constitutes a self-contained statistical evaluation against internal benchmarks rather than any self-referential loop. No equations or claims reduce by construction to fitted inputs renamed as predictions, no load-bearing self-citations appear, and no uniqueness theorems or ansatzes are smuggled in. The reported improvements and calibrated margins are falsifiable via the hold-out procedure and do not collapse into tautology.

Axiom & Free-Parameter Ledger

3 free parameters · 2 axioms · 0 invented entities

The central claim rests on the continued validity of the log-distance multi-wall baseline, the assumption that environmental covariates act as additive predictors, and multiple regression coefficients fitted directly to the campaign data.

free parameters (3)
  • regression coefficients for environmental covariates
    Fitted by regularized MLR, Bayesian LR, and selective quadratic MLR to the 12-month dataset.
  • parameters of the 3-component Gaussian mixture
    Fitted to the out-of-fold residuals to capture the observed non-Gaussian shape.
  • upper-tail percentile threshold for fade margin
    Chosen to target 1% outage and then validated on the hold-out set.
axioms (2)
  • domain assumption Log-distance multi-wall model remains an appropriate baseline once environmental covariates are added
    The framework is explicitly described as augmenting this baseline.
  • domain assumption Environmental measurements are sufficiently independent of distance and wall losses to be treated as separate regressors
    Implicit in the inclusion of humidity, temperature, CO2, PM, and pressure as predictors.

pith-pipeline@v0.9.0 · 5871 in / 1651 out tokens · 51216 ms · 2026-05-18T10:15:06.376953+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

68 extracted references · 68 canonical work pages

  1. [1]

    Journal of Sensor and Actuator Networks6(2), 7 (2017) https://doi.org/10.3390/ jsan6020007

    Cattani, M., Boano, C.A., Römer, K.: An Experimental Evaluation of the Reliability of LoRa Long-Range Low- Power Wireless Communication. Journal of Sensor and Actuator Networks6(2), 7 (2017) https://doi.org/10.3390/ jsan6020007

    work page 2017

  2. [2]

    Pervasive and Mobile Computing84, 101640 (2022) https://doi.org/10.1016/j.pmcj.2022.101640

    Grübel, J., Thrash, T., Aguilar, L., Gath-Morad, M., Hélal, D., Sumner, R.W., Hölscher, C., Schinazi, V.R.: Dense Indoor Sensor Networks: Towards passively sensing human presence with LoRaWAN. Pervasive and Mobile Computing84, 101640 (2022) https://doi.org/10.1016/j.pmcj.2022.101640

    work page doi:10.1016/j.pmcj.2022.101640 2022

  3. [3]

    Data8(1), 4 (2023) https://doi.org/10

    González-Palacio, M., Tobón-Vallejo, D., Sepúlveda-Cano, L.M., Rúa, S., Pau, G., Le, L.B.: LoRaWAN Path Loss Measurements in an Urban Scenario including Environmental Effects. Data8(1), 4 (2023) https://doi.org/10. 3390/data8010004

    work page 2023

  4. [4]

    Computers14(1), 15 (2025) https://doi.org/10.3390/computers14010015

    Siddiky, M.N.A., Rahman, M.E., Uzzal, M.S., Kabir, H.M.D.: A Comprehensive Exploration of 6G Wireless Communication Technologies. Computers14(1), 15 (2025) https://doi.org/10.3390/computers14010015

    work page doi:10.3390/computers14010015 2025

  5. [5]

    In: 2024 IEEE 25th International Symposium on a World of Wireless, Mobile and Multimedia Networks (WoWMoM), pp

    Szafranski, D.: Predictability of LoRaWAN Link Quality based on Weather Data: Insights from a Long-Term Study. In: 2024 IEEE 25th International Symposium on a World of Wireless, Mobile and Multimedia Networks (WoWMoM), pp. 249–258 (2024). https://doi.org/10.1109/WoWMoM60985.2024.00048

    work page doi:10.1109/wowmom60985.2024.00048 2024

  6. [6]

    IEEE Access13, 83148–83170 (2025) https://doi.org/ 10.1109/ACCESS.2025.3569164 20

    Obiri, N.M., van Laerhoven, K.: A Comprehensive Data Description for LoRaWAN Path Loss Measurements in an Indoor Office Setting: Effects of Environmental Factors. IEEE Access13, 83148–83170 (2025) https://doi.org/ 10.1109/ACCESS.2025.3569164 20

    work page doi:10.1109/access.2025.3569164 2025

  7. [7]

    In: 2025 Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit), pp

    Obiri, N.M., Van Laerhoven, K.: A Statistical Evaluation of Indoor LoRaWAN Environment-Aware Propagation for 6G: MLR, ANOVA, and Residual Distribution Analysis. In: 2025 Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit), pp. 494–499 (2025). https://doi.org/10.1109/EuCNC/ 6GSummit63408.2025.11037092

    work page doi:10.1109/eucnc/ 2025

  8. [8]

    IEEE Internet of Things Journal10(12), 10725–10739 (2023) https://doi.org/10.1109/JIOT.2023.3239827

    González-Palacio, M., Tobón-Vallejo, D., Sepúlveda-Cano, L.M., Rúa, S., Le, L.B.: Machine-Learning-Based Com- bined Path Loss and Shadowing Model in LoRaWAN for Energy Efficiency Enhancement. IEEE Internet of Things Journal10(12), 10725–10739 (2023) https://doi.org/10.1109/JIOT.2023.3239827

    work page doi:10.1109/jiot.2023.3239827 2023

  9. [9]

    IEEE Open Journal of the Communications Society5, 6713–6735 (2024) https://doi.org/10.1109/OJCOMS.2024.3484002

    Obiri, N.M., Van Laerhoven, K.: A Survey of LoRaWAN-Integrated Wearable Sensor Networks for Human Activ- ity Recognition: Applications, Challenges and Possible Solutions. IEEE Open Journal of the Communications Society5, 6713–6735 (2024) https://doi.org/10.1109/OJCOMS.2024.3484002

    work page doi:10.1109/ojcoms.2024.3484002 2024

  10. [10]

    Sensors 24(12), 3877 (2024) https://doi.org/10.3390/s24123877

    Azevedo, J.A., Mendonça, F.: A Critical Review of the Propagation Models Employed in LoRa Systems. Sensors 24(12), 3877 (2024) https://doi.org/10.3390/s24123877

    work page doi:10.3390/s24123877 2024

  11. [11]

    Sensors23(6), 3283 (2023) https://doi.org/10.3390/s23063283

    Robles-Enciso, R., Morales-Aragón, I.P., Serna-Sabater, A., Martínez-Inglés, M.T., Mateo-Aroca, A., Molina- Garcia-Pardo, J.-M., Juan-Llácer, L.: LoRa, Zigbee and 5G Propagation and Transmission Performance in an Indoor Environment at 868 MHz. Sensors23(6), 3283 (2023) https://doi.org/10.3390/s23063283

    work page doi:10.3390/s23063283 2023

  12. [12]

    In: 2023 6th International Seminar on Research of Information Technology and Intelligent Systems (ISRITI), pp

    Anisah, I., Wirawan, Suwadi, Yuliana, M.: Experimental Results of LoRa Network Radio Propagation Modeling in Campus Area. In: 2023 6th International Seminar on Research of Information Technology and Intelligent Systems (ISRITI), pp. 427–432 (2023). https://doi.org/10.1109/ISRITI60336.2023.10467989

    work page doi:10.1109/isriti60336.2023.10467989 2023

  13. [13]

    In: 2019 13th European Conference on Antennas and Propagation (EuCAP), pp

    Bertoldo, S., Paredes, M., Carosso, L., Allegretti, M., Savi, P.: Empirical indoor propagation models for LoRa radio link in an office environment. In: 2019 13th European Conference on Antennas and Propagation (EuCAP), pp. 1–5 (2019)

    work page 2019

  14. [14]

    Technical report, International Telecommu- nication Union, Geneva, Switzerland (2021)

    Propagation data and prediction methods for the planning of indoor radiocommunication systems and radio local area networks in the frequency range 300 MHz to 450 GHz. Technical report, International Telecommu- nication Union, Geneva, Switzerland (2021)

    work page 2021

  15. [15]

    Final Report, European Commission, Directorate- General for the Information Society and Media, Luxembourg (1999)

    Digital mobile radio towards future generation systems. Final Report, European Commission, Directorate- General for the Information Society and Media, Luxembourg (1999)

    work page 1999

  16. [16]

    In: SBMO/IEEE MTT-S International Conference on Microwave and Optoelectronics, 2005., pp

    Lima, A.G.M., Menezes, L.F.: Motley-Keenan model adjusted to the thickness of the wall. In: SBMO/IEEE MTT-S International Conference on Microwave and Optoelectronics, 2005., pp. 180–182 (2005). https://doi.org/10.1109/ IMOC.2005.1580040

    work page arXiv 2005

  17. [17]

    Communications and Network17(1), 1–19 (2025) https://doi.org/10.4236/cn.2025.171001

    Zhong, C.: Measurement and Modeling of LoRa Signal in Multi-Floor Home Environment. Communications and Network17(1), 1–19 (2025) https://doi.org/10.4236/cn.2025.171001

    work page doi:10.4236/cn.2025.171001 2025

  18. [18]

    In: 2023 IEEE 11th Interna- tional Conference on Systems and Control (ICSC), pp

    Alkhazmi, E.H., Elkawafi, S.M., Aldarrat, A.A., Abbas, M.A., Abubakr, H., Shamatah, H.A.: Analysis of Real- World LoRaWAN Network Performance Across Outdoor and Indoor Scenarios. In: 2023 IEEE 11th Interna- tional Conference on Systems and Control (ICSC), pp. 329–334 (2023). https://doi.org/10.1109/ICSC58660.2023. 10449775

    work page doi:10.1109/icsc58660.2023 2023

  19. [19]

    In: 2021 6th International Conference for Convergence in Technology (I2CT), pp

    Muppala, R., Navnit, A., Poondla, S., Hussain, A.M.: Investigation of Indoor LoRaWAN Signal Propagation for Real-World Applications. In: 2021 6th International Conference for Convergence in Technology (I2CT), pp. 1–5 (2021). https://doi.org/10.1109/I2CT51068.2021.9418173

    work page doi:10.1109/i2ct51068.2021.9418173 2021

  20. [20]

    Computers11(2), 25 (2022) https://doi.org/10.3390/computers11020025 21

    Harinda, E., Wixted, A.J., Qureshi, A.-U.-H., Larijani, H., Gibson, R.M.: Performance of a Live Multi-Gateway LoRaWAN and Interference Measurement across Indoor and Outdoor Localities. Computers11(2), 25 (2022) https://doi.org/10.3390/computers11020025 21

    work page doi:10.3390/computers11020025 2022

  21. [21]

    IEEE Access6, 30149– 30161 (2018) https://doi.org/10.1109/ACCESS.2018.2843325

    Sadowski, S., Spachos, P.: RSSI-Based Indoor Localization With the Internet of Things. IEEE Access6, 30149– 30161 (2018) https://doi.org/10.1109/ACCESS.2018.2843325

    work page doi:10.1109/access.2018.2843325 2018

  22. [22]

    Computer Networks243, 110258 (2024) https://doi.org/10.1016/j.comnet.2024.110258

    Guerra, R.R., Vizziello, A., Savazzi, P., Goldoni, E., Gamba, P.: Forecasting LoRaWAN RSSI using weather param- eters: A comparative study of ARIMA, artificial intelligence and hybrid approaches. Computer Networks243, 110258 (2024) https://doi.org/10.1016/j.comnet.2024.110258

    work page doi:10.1016/j.comnet.2024.110258 2024

  23. [23]

    Wireless Personal Communications134(1), 339–360 (2024) https://doi.org/10.1007/s11277-024-10911-z

    Aksoy, A., Yıldız, Ö., Karlık, S.E.: Comparative Analysis of End Device and Field Test Device Measure- ments for RSSI, SNR and SF Performance Parameters in an Indoor LoRaWAN Network. Wireless Personal Communications134(1), 339–360 (2024) https://doi.org/10.1007/s11277-024-10911-z

    work page doi:10.1007/s11277-024-10911-z 2024

  24. [24]

    IEEE Access12, 83205–83216 (2024) https://doi.org/10.1109/ACCESS.2024.3412849

    Vo, H., Hoang Long Nguyen, V., Tran, V.L., Ferrero, F., Lee, F.-Y., Tsai, M.-H.: Advance Path Loss Model for Distance Estimation Using LoRaWAN Network’s Received Signal Strength Indicator (RSSI). IEEE Access12, 83205–83216 (2024) https://doi.org/10.1109/ACCESS.2024.3412849

    work page doi:10.1109/access.2024.3412849 2024

  25. [25]

    University of Bath, Bath, UK (2002)

    Faraway, J.J.: Practical Regression and Anova using R. University of Bath, Bath, UK (2002)

    work page 2002

  26. [26]

    O’Reilly Media, Inc., Sebastopol, CA (2019)

    Géron, A.: Hands-on Machine Learning with Scikit-Learn, Keras, And TensorFlow, 2nd edition (updated for tensorflow 2) edn. O’Reilly Media, Inc., Sebastopol, CA (2019)

    work page 2019

  27. [27]

    IEEE Transactions on Mobile Computing 16(8), 2079–2092 (2017) https://doi.org/10.1109/TMC.2016.2616465

    Khalajmehrabadi, A., Gatsis, N., Pack, D.J., Akopian, D.: A Joint Indoor WLAN Localization and Outlier Detec- tion Scheme Using LASSO and Elastic-Net Optimization Techniques. IEEE Transactions on Mobile Computing 16(8), 2079–2092 (2017) https://doi.org/10.1109/TMC.2016.2616465

    work page doi:10.1109/tmc.2016.2616465 2079

  28. [28]

    In: 2024 IEEE International Conference on Smart Power Control and Renewable Energy (ICSPCRE), pp

    Bhavanam, B.P.R., Ragam, P.: Exploring LoRa Signal Propagation in Indoor and Outdoor Environments: A Comparative Study. In: 2024 IEEE International Conference on Smart Power Control and Renewable Energy (ICSPCRE), pp. 1–6 (2024). https://doi.org/10.1109/ICSPCRE62303.2024.10675046

    work page doi:10.1109/icspcre62303.2024.10675046 2024

  29. [29]

    IEEE Internet of Things Journal12(12), 20251–20260 (2025) https://doi.org/10.1109/JIOT.2025.3542370

    Lin, B., Chen, J., Xu, B., Chao, J., Zheng, B., Pang, G., Luo, J., Ghassemlooy, Z.: Indoor NLOS-VLP System Based on Image Sensor and Pixel Coordinate Fingerprinting. IEEE Internet of Things Journal12(12), 20251–20260 (2025) https://doi.org/10.1109/JIOT.2025.3542370

    work page doi:10.1109/jiot.2025.3542370 2025

  30. [30]

    Journal of the American Statistical Association103(481), 410–423 (2008) https://doi.org/10.1198/ 016214507000001337

    Liang, F., Paulo, R., Molina, G., Clyde, M.A., Berger, J.O.: Mixtures of g Priors for Bayesian Variable Selection. Journal of the American Statistical Association103(481), 410–423 (2008) https://doi.org/10.1198/ 016214507000001337

    work page 2008

  31. [31]

    CRC Press, New York (2013)

    Gelman, A., Carlin, J.B., Stern, H.S., Dunson, D.B., Vehtari, A., Rubin, D.B.: Bayesian Data Analysis, Third Edition. CRC Press, New York (2013)

    work page 2013

  32. [32]

    Statistics and Computing27(5), 1413–1432 (2017) https://doi.org/10.1007/s11222-016-9696-4

    Vehtari, A., Gelman, A., Gabry, J.: Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC. Statistics and Computing27(5), 1413–1432 (2017) https://doi.org/10.1007/s11222-016-9696-4

    work page doi:10.1007/s11222-016-9696-4 2017

  33. [33]

    Heliyon9(9), 19685 (2023) https://doi.org/10.1016/j.heliyon.2023.e19685

    Elmezughi, M.K., Salih, O., Afullo, T.J., Duffy, K.J.: Path loss modeling based on neural networks and ensemble method for future wireless networks. Heliyon9(9), 19685 (2023) https://doi.org/10.1016/j.heliyon.2023.e19685

    work page doi:10.1016/j.heliyon.2023.e19685 2023

  34. [34]

    Sensors22(13), 4967 (2022) https://doi.org/10.3390/s22134967

    Elmezughi, M.K., Salih, O., Afullo, T.J., Duffy, K.J.: Comparative Analysis of Major Machine-Learning-Based Path Loss Models for Enclosed Indoor Channels. Sensors22(13), 4967 (2022) https://doi.org/10.3390/s22134967

    work page doi:10.3390/s22134967 2022

  35. [35]

    Sensors (Basel, Switzerland)24(3), 860 (2024) https://doi.org/10.3390/ s24030860

    Hosseinzadeh, S., Ashawa, M., Owoh, N., Larijani, H., Curtis, K.: Explainable Machine Learning for LoRaWAN Link Budget Analysis and Modeling. Sensors (Basel, Switzerland)24(3), 860 (2024) https://doi.org/10.3390/ s24030860

    work page 2024

  36. [36]

    Sensors25(13), 4101 (2025) https://doi.org/10.3390/s25134101

    Alkhayyal, M.A., Mostafa, A.M.: Enhancing LoRaWAN Performance Using Boosting Machine Learning Algo- rithms Under Environmental Variations. Sensors25(13), 4101 (2025) https://doi.org/10.3390/s25134101

    work page doi:10.3390/s25134101 2025

  37. [37]

    John Wiley & Sons, Hoboken, New Jersey (2017) 22

    Montgomery, D.C.: Design and Analysis of Experiments. John Wiley & Sons, Hoboken, New Jersey (2017) 22

    work page 2017

  38. [38]

    Statistics and Computing13(2), 163–167 (2003) https://doi.org/10.1023/A:1023260610025

    Langsrud, Ø.: ANOVA for unbalanced data: Use Type II instead of Type III sums of squares. Statistics and Computing13(2), 163–167 (2003) https://doi.org/10.1023/A:1023260610025

    work page doi:10.1023/a:1023260610025 2003

  39. [39]

    International Journal of Ad hoc, Sensor & Ubiquitous Computing5(6), 21–29 (2014) https://doi.org/10.5121/ ijasuc.2014.5603 arXiv:1501.01073 [cs]

    Chelloug, S.A.: Impact of the Temperature and Humidity Variations on Link Quality of xm1000 Mote Sensors. International Journal of Ad hoc, Sensor & Ubiquitous Computing5(6), 21–29 (2014) https://doi.org/10.5121/ ijasuc.2014.5603 arXiv:1501.01073 [cs]

    work page arXiv 2014

  40. [40]

    In: 8th International Symposium on Modeling and Optimization in Mobile, Ad Hoc, And Wireless Networks, pp

    Christmann, D., Martinovic, I.: Experimental design and analysis of transmission properties in an indoor wire- less sensor network. In: 8th International Symposium on Modeling and Optimization in Mobile, Ad Hoc, And Wireless Networks, pp. 342–347 (2010)

    work page 2010

  41. [41]

    IEEE Sensors Journal15(10), 5483–5493 (2015) https://doi.org/10.1109/JSEN.2015.2443380

    Harb, H., Makhoul, A., Couturier, R.: An Enhanced K-Means and ANOVA-Based Clustering Approach for Sim- ilarity Aggregation in Underwater Wireless Sensor Networks. IEEE Sensors Journal15(10), 5483–5493 (2015) https://doi.org/10.1109/JSEN.2015.2443380

    work page doi:10.1109/jsen.2015.2443380 2015

  42. [42]

    IET Microwaves, Antennas & Propagation11(9), 1203–1211 (2017) https://doi.org/10.1049/iet-map.2016.0416

    Allen, B., Mahato, S., Gao, Y., Salous, S.: Indoor-to-outdoor empirical path loss modelling for femtocell networks at 0.9, 2, 2.5 and 3.5 GHz using singular value decomposition. IET Microwaves, Antennas & Propagation11(9), 1203–1211 (2017) https://doi.org/10.1049/iet-map.2016.0416

    work page doi:10.1049/iet-map.2016.0416 2017

  43. [43]

    Sensors20(14), 3941 (2020) https://doi.org/10.3390/s20143941

    Wang, Y., Ren, W., Cheng, L., Zou, J.: A Grey Model and Mixture Gaussian Residual Analysis-Based Position Estimator in an Indoor Environment. Sensors20(14), 3941 (2020) https://doi.org/10.3390/s20143941

    work page doi:10.3390/s20143941 2020

  44. [44]

    Routledge, New York (2018)

    Silverman, B.W.: Density Estimation for Statistics and Data Analysis. Routledge, New York (2018). https://doi. org/10.1201/9781315140919

    work page doi:10.1201/9781315140919 2018

  45. [45]

    Practical Assessment, Research & Evaluation24(1) (2019)

    Astivia, O.L.O., Zumbo, B.D.: Heteroskedasticity in Multiple Regression Analysis: What it is, How to Detect it and How to Solve it with Applications in R and SPSS. Practical Assessment, Research & Evaluation24(1) (2019)

    work page 2019

  46. [46]

    Physical Therapy77(12), 1755–1761 (1997) https://doi.org/10.1093/ptj/77.12.1755

    Chan, Y., Walmsley, R.P.: Learning and Understanding the Kruskal-Wallis One-Way Analysis-of-Variance-by- Ranks Test for Differences Among Three or More Independent Groups. Physical Therapy77(12), 1755–1761 (1997) https://doi.org/10.1093/ptj/77.12.1755

    work page doi:10.1093/ptj/77.12.1755 1997

  47. [47]

    IEEE Transactions on Antennas and Propagation69(11), 7782–7794 (2021) https://doi.org/10.1109/TAP.2021.3076171

    Zhang, L., Cotton, S.L., Yoo, S.K., Ngo, H.Q., Fernández, M., Scanlon, W.G.: A Time Series-Based Study of Corre- lation, Channel Power Imbalance, and Diversity Gain in Indoor Distributed Antenna Systems at 60 GHz. IEEE Transactions on Antennas and Propagation69(11), 7782–7794 (2021) https://doi.org/10.1109/TAP.2021.3076171

    work page doi:10.1109/tap.2021.3076171 2021

  48. [48]

    Sasi Kiran Gaddipati et al

    Efron, B.: Bootstrap Methods: Another Look at the Jackknife. The Annals of Statistics7(1), 1–26 (1979) https: //doi.org/10.1214/aos/1176344552

    work page doi:10.1214/aos/1176344552 1979

  49. [49]

    In: 2017 International Conference on Internet of Things for the Global Community (IoTGC), pp

    Ayele, E.D., Hakkenberg, C., Meijers, J.P., Zhang, K., Meratnia, N., Havinga, P.J.M.: Performance analysis of LoRa radio for an indoor IoT applications. In: 2017 International Conference on Internet of Things for the Global Community (IoTGC), pp. 1–8 (2017). https://doi.org/10.1109/IoTGC.2017.8008973

    work page doi:10.1109/iotgc.2017.8008973 2017

  50. [50]

    Big Data and Cognitive Computing1(1), 7 (2017) https: //doi.org/10.3390/bdcc1010007

    Hosseinzadeh, S., Almoathen, M., Larijani, H., Curtis, K.: A Neural Network Propagation Model for LoRaWAN and Critical Analysis with Real-World Measurements. Big Data and Cognitive Computing1(1), 7 (2017) https: //doi.org/10.3390/bdcc1010007

    work page doi:10.3390/bdcc1010007 2017

  51. [51]

    International Journal of Wireless Information Networks 24(2), 153–165 (2017) https://doi.org/10.1007/s10776-017-0341-8

    Petäjäjärvi, J., Mikhaylov, K., Yasmin, R., Hämäläinen, M., Iinatti, J.: Evaluation of LoRa LPWAN Technology for Indoor Remote Health and Wellbeing Monitoring. International Journal of Wireless Information Networks 24(2), 153–165 (2017) https://doi.org/10.1007/s10776-017-0341-8

    work page doi:10.1007/s10776-017-0341-8 2017

  52. [52]

    In: 2018 3rd International Conference on Computer and Communication Systems (ICCCS), pp

    Erbati, M.M., Schiele, G., Batke, G.: Analysis of LoRaWAN technology in an Outdoor and an Indoor Scenario in Duisburg-Germany. In: 2018 3rd International Conference on Computer and Communication Systems (ICCCS), pp. 273–277 (2018). https://doi.org/10.1109/CCOMS.2018.8463224 23

    work page doi:10.1109/ccoms.2018.8463224 2018

  53. [53]

    IEEE Internet of Things Journal6(2), 2366–2378 (2019) https: //doi.org/10.1109/JIOT.2019.2906838

    El Chall, R., Lahoud, S., El Helou, M.: LoRaWAN Network: Radio Propagation Models and Performance Eval- uation in Various Environments in Lebanon. IEEE Internet of Things Journal6(2), 2366–2378 (2019) https: //doi.org/10.1109/JIOT.2019.2906838

    work page doi:10.1109/jiot.2019.2906838 2019

  54. [54]

    In: 2019 IEEE 9th Symposium on Computer Applications & Industrial Electronics (ISCAIE), pp

    Muzammir, M.I., Abidin, H.Z., Abdullah, S.A.C., Zaman, F.H.K.: Performance Analysis of LoRaWAN for Indoor Application. In: 2019 IEEE 9th Symposium on Computer Applications & Industrial Electronics (ISCAIE), pp. 156–159 (2019). https://doi.org/10.1109/ISCAIE.2019.8743982

    work page doi:10.1109/iscaie.2019.8743982 2019

  55. [55]

    IFAC-PapersOnLine54(4), 159–164 (2021) https://doi.org/10.1016/j.ifacol.2021.10.027

    Saban, M., Aghzout, O., Medus, L.D., Rosado, A.: Experimental Analysis of IoT Networks Based on LoRa/Lo- RaWAN under Indoor and Outdoor Environments: Performance and Limitations. IFAC-PapersOnLine54(4), 159–164 (2021) https://doi.org/10.1016/j.ifacol.2021.10.027

    work page doi:10.1016/j.ifacol.2021.10.027 2021

  56. [56]

    Cambridge University Press, Cambridge (2014)

    Azzalini, A., Capitanio, A.: The Skew-Normal and Related Families. Cambridge University Press, Cambridge (2014)

    work page 2014

  57. [57]

    John Wiley & Sons, New York, USA (1995)

    Johnson, N.L., Kotz, S., Balakrishnan, N.: Continuous Univariate Distributions, Volume 2. John Wiley & Sons, New York, USA (1995)

    work page 1995

  58. [58]

    In: Li, S.Z., Jain, A

    Reynolds, D.: Gaussian Mixture Models. In: Li, S.Z., Jain, A. (eds.) Encyclopedia of Biometrics, pp. 659–663. Springer, Boston, MA (2009). https://doi.org/10.1007/978-0-387-73003-5_196

    work page doi:10.1007/978-0-387-73003-5_196 2009

  59. [59]

    Proceedings of the IRE34(5), 254–256 (1946) https://doi

    Friis, H.T.: A Note on a Simple Transmission Formula. Proceedings of the IRE34(5), 254–256 (1946) https://doi. org/10.1109/JRPROC.1946.234568

    work page doi:10.1109/jrproc.1946.234568 1946

  60. [60]

    IEEE Transactions on Automatic Control , volume =

    Akaike, H.: A new look at the statistical model identification. IEEE Transactions on Automatic Control19(6), 716–723 (1974) https://doi.org/10.1109/TAC.1974.1100705

    work page doi:10.1109/tac.1974.1100705 1974

  61. [61]

    Sasi Kiran Gaddipati et al

    Schwarz, G.: Estimating the Dimension of a Model. The Annals of Statistics6(2), 461–464 (1978) https://doi.org/ 10.1214/aos/1176344136

    work page doi:10.1214/aos/1176344136 1978

  62. [62]

    Journal of the American Statistical Association46(253), 68–78 (1951) https://doi.org/10.1080/01621459.1951.10500769

    Massey Jr., F.J.: The Kolmogorov-Smirnov Test for Goodness of Fit. Journal of the American Statistical Association46(253), 68–78 (1951) https://doi.org/10.1080/01621459.1951.10500769

    work page doi:10.1080/01621459.1951.10500769 1951

  63. [63]

    Scientific Reports13(1), 6385 (2023) https://doi.org/10.1038/s41598-023-33598-x

    Papasotiriou, E.N., Boulogeorgos, A.-A.A., Alexiou, A.: Outdoor THz fading modeling by means of gaussian and gamma mixture distributions. Scientific Reports13(1), 6385 (2023) https://doi.org/10.1038/s41598-023-33598-x

    work page doi:10.1038/s41598-023-33598-x 2023

  64. [64]

    Cambridge University Press, Cambridge (2005)

    Goldsmith, A.: Wireless Communications. Cambridge University Press, Cambridge (2005)

    work page 2005

  65. [65]

    Final Report 2604/BMEM/R/3/2.0, Ofcom, United Kingdom (2014)

    Rudd, R., Craig, K., Ganley, M., Hartless, R.: Building Materials and Propagation. Final Report 2604/BMEM/R/3/2.0, Ofcom, United Kingdom (2014)

    work page 2014

  66. [66]

    Wang, F., Lv, Y., Zhu, M., Ding, H., Han, J.: XRF55: A Radio Frequency Dataset for Human Indoor Action Analysis. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.8(1), 21–12134 (2024) https://doi.org/10.1145/3643543

    work page doi:10.1145/3643543 2024

  67. [67]

    Elsevier, Amsterdam, Netherlands (2010)

    Sebastian, M.T.: Dielectric Materials for Wireless Communication. Elsevier, Amsterdam, Netherlands (2010)

    work page 2010

  68. [68]

    Prentice Hall communications engineering and emerging technologies series

    Rappaport, T.S.: Wireless Communications: Principles and Practice. Prentice Hall communications engineering and emerging technologies series. Prentice Hall, Upper Saddle River, N.J (2002) 24

    work page 2002