Gateway-level Early Warning of LoRaWAN Link Degradation Using LoED Measurements
DOI:
https://doi.org/10.26636/jtit.2026.4.2789Keywords:
early warning, edge monitoring, link degradation, LoRaWANAbstract
LoRaWAN gateways observe packet-level metadata that may support reliability monitoring without modifying end devices or the protocol stack. We formulate short-term link degradation as a next-window warning task and aggregate LoED records into five-minute gateway windows described by CRC success rate, RSSI, SNR, spreading factor, traffic, frequency diversity, and historical features. A degradation event is a 10% or 20% decrease in the following window's CRC success rate. Class-balanced logistic regression (LR), random forest, XGBoost, and rule baselines are evaluated chronologically and with leave-one-gateway-out (LOGO) testing. LR obtains chronological PR-AUC values of 0.563 and 0.367 and exceeds both non-linear models in date-clustered randomization tests after Holm correction. Its LOGO means are also the highest, but the differences are not statistically significant, as only seven and five gateways contain both classes. Removing current CRC alone has a negligible effect, whereas removing all CRC history substantially reduces performance. Performance is further characterized across four window durations, probability calibration schemes, and explicit sequence models. A strictly chronological offline replay processes held-out windows with P99 scoring latency below 15 ms. The results support risk ranking within the LoED deployment, while gateway heterogeneity and calibration drift limit broader transfer claims.
Downloads
References
[1] LoRa Alliance, "LoRaWAN L2 1.0.4 Specification", Technical Specification, 2020.
View in Google Scholar
[2] Semtech Corporation, "LoRa and LoRaWAN: A Technical Overview", Technical White Paper, 2024 (https://www.semtech.com/uploads/technology/LoRa/lora-and-lorawan.pdf).
View in Google Scholar
[3] F. Adelantado et al., "Understanding the Limits of LoRaWAN", IEEE Communications Magazine, vol. 55, pp. 34-40, 2017. DOI: https://doi.org/10.1109/MCOM.2017.1600613
View in Google Scholar
[4] N. Blenn and F. Kuipers, "LoRaWAN in the Wild: Measurements from The Things Network", arXiv, 2017.
View in Google Scholar
[5] S. Li, U. Raza, and A. Khan, "How Agile is the Adaptive Data Rate Mechanism of LoRaWAN?", 2018 IEEE Global Communications Conference, Abu Dhabi, UAE, 2018. DOI: https://doi.org/10.1109/GLOCOM.2018.8647469
View in Google Scholar
[6] J. Finnegan, R. Farrell, and S. Brown, "Analysis and Enhancement of the LoRaWAN Adaptive Data Rate Scheme", IEEE Internet of Things Journal, vol. 7, pp. 7171-7180, 2020. DOI: https://doi.org/10.1109/JIOT.2020.2982745
View in Google Scholar
[7] A. Hoeller, R.D. Souza, S. Montejo-Sanchez, and H. Alves, "Performance Analysis of Single-cell Adaptive Data Rate-enabled LoRaWAN", IEEE Wireless Communications Letters, vol. 9, pp. 911-914, 2020. DOI: https://doi.org/10.1109/LWC.2020.2975604
View in Google Scholar
[8] L. Bhatia, M. Breza, R. Marfievici, and J.A. McCann, "LoED: The LoRaWAN at the Edge Dataset", Proc. of 3rd Workshop on Data Acquisition to Analysis, pp. 7-8, 2020. DOI: https://doi.org/10.1145/3419016.3431491
View in Google Scholar
[9] P. Spadaccino, F.G. Crino, and F. Cuomo, "LoRaWAN Behaviour Analysis Through Dataset Traffic Investigation", Sensors, vol. 22, art. no. 2470, 2022. DOI: https://doi.org/10.3390/s22072470
View in Google Scholar
[10] G. Cerar, H. Yetgin, M. Mohorcic, and C. Fortuna, "Machine Learning for Wireless Link Quality Estimation: A Survey", IEEE Communications Surveys & Tutorials, vol. 23, pp. 696-728, 2021. DOI: https://doi.org/10.1109/COMST.2021.3053615
View in Google Scholar
[11] G. Cerar, H. Yetgin, M. Mohorcic, and C. Fortuna, "On Designing a Machine Learning Based Wireless Link Quality Classifier", 2020 IEEE 31st Annual International Symposium on Personal, Indoor and Mobile Radio Communications, London, UK, 2020. DOI: https://doi.org/10.1109/PIMRC48278.2020.9217171
View in Google Scholar
[12] Y. Kanto and K. Watabe, "Wireless Link Quality Estimation Using LSTM Model", 2024 IEEE/IFIP Network Operations and Management Symposium (NOMS), Seoul, Republic of Korea, 2024. DOI: https://doi.org/10.1109/NOMS59830.2024.10575638
View in Google Scholar
[13] A. Prakash, N. Choudhury, A. Hazarika, and A. Gorrela, "Effective Feature Selection for Predicting Spreading Factor with ML in Large LoRaWAN-based Mobile IoT Networks", 2025 National Conference on Communications, New Delhi, India, 2025. DOI: https://doi.org/10.1109/NCC63735.2025.10983488
View in Google Scholar
[14] P. Maurya et al., "A Comprehensive Survey of Data-driven Solutions for LoRaWAN: Challenges and Future Directions", ACM Transactions on Internet of Things, vol. 6, art. no. 6, 2025. DOI: https://doi.org/10.1145/3711953
View in Google Scholar
[15] S. Hosseinzadeh et al., "Explainable Machine Learning for LoRaWAN Link Budget Analysis and Modeling", Sensors, vol. 24, art. no. 860, 2024. DOI: https://doi.org/10.3390/s24030860
View in Google Scholar
[16] J. Park, S. Samarakoon, M. Bennis, and M. Debbah, "Wireless Network Intelligence at the Edge", Proceedings of the IEEE, vol. 107, pp. 2204-2239, 2019. DOI: https://doi.org/10.1109/JPROC.2019.2941458
View in Google Scholar
[17] B.J. Bashar and H. Abdullah, "Analyzing Performance of Eigenvalue-based Spectrum Sensing within LoRaCog Framework", Journal of Telecommunications and Information Technology, no. 2, pp. 67-74, 2026. DOI: https://doi.org/10.26636/jtit.2026.2.2576
View in Google Scholar
[18] T. Chen and C. Guestrin, "XGBoost: A Scalable Tree Boosting System", Proc. of 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 785-794, 2016. DOI: https://doi.org/10.1145/2939672.2939785
View in Google Scholar
[19] T. Saito and M. Rehmsmeier, "The Precision-recall Plot Is More Informative than the ROC Plot when Evaluating Binary Classifiers on Imbalanced Datasets", PLOS ONE, vol. 10, art. no. e0118432, 2015. DOI: https://doi.org/10.1371/journal.pone.0118432
View in Google Scholar
[20] S. Holm, "A Simple Sequentially Rejective Multiple Test Procedure", Scandinavian Journal of Statistics, vol. 6, pp. 65-70, 1979.
View in Google Scholar
[21] N. Cliff, "Dominance Statistics: Ordinal Analyses to Answer Ordinal Questions", Psychological Bulletin, vol. 114, pp. 494-509, 1993. DOI: https://doi.org/10.1037/0033-2909.114.3.494
View in Google Scholar
[22] C. Guo, G. Pleiss, Y. Sun, and K.Q. Weinberger, "On Calibration of Modern Neural Networks", arXiv, 2017.
View in Google Scholar
[23] K. Cho et al., "Learning Phrase Representations using RNN Encoder-decoder for Statistical Machine Translation", Proc. of the 2014 Conference on Empirical Methods in Natural Language Processing, pp. 1724-1734, 2014. DOI: https://doi.org/10.3115/v1/D14-1179
View in Google Scholar
[24] G.W. Brier, "Verification of Forecasts Expressed in Terms of Probability", Monthly Weather Review, vol. 78, pp. 1-3, 1950. DOI: https://doi.org/10.1175/1520-0493(1950)078<0001:VOFEIT>2.0.CO;2
View in Google Scholar
[25] A.C. Davison and D.V. Hinkley, Bootstrap Methods and Their Application, Cambridge University Press, 1997. DOI: https://doi.org/10.1017/CBO9780511802843
View in Google Scholar
Downloads
Submitted
Published
Issue
Section
License
Copyright (c) 2026 Esma Fazilet Karagülle, Faruk Baturalp Günay, Esra Odabaş Yıldırım
This work is licensed under a Creative Commons Attribution 4.0 International License.
