aariflhoussaine/indoor-lorawan-environment-classification

Partial dataset (8,774 packets) and analysis code (MATLAB + Python) for indoor LoRaWAN propagation characterization and ML-based environment classification across 4 environments in Casablanca, Morocco. # Indoor LoRaWAN Multi-Environment Dataset and Analysis Code (Partial Release) This repository accompanies the study: > Aarif, L., Ahaidous, K., Tabaa, M., Hachimi, H., Monteiro, F. *Empirical > Characterization and Machine Learning-Based Classification of Indoor > LoRaWAN Propagation Environments.* It provides a partial sample of the raw measurement dataset and the analysis code used to produce the empirical radio link characterization results (Section 4.1 of the paper). ## Contents - **`data/`** — 7 CSV files covering the four indoor environments (Residential, Laboratory, Multi-floor building, Underground Industrial), 8,774 packets total (a ~20% sample of the full 43,100-packet dataset). See `data/README.md` for column descriptions and the environment/condition mapping. The `data/pdr_raw/` subfolder is reserved for per-node raw files used specifically for the Packet Delivery Rate calculation (Table 7) — see that folder's README. - **`matlab/`** — MATLAB scripts reproducing the RSSI, SNR, frequency, and correlation analyses (Figures 10–13). See `matlab/README.md` for the principle behind each analysis and a list of open methodological points still to be confirmed. - **`code/`** — Python notebook for the supervised machine learning environment classification pipeline (Section 4.2 of the paper: feature engineering, 18-classifier comparison, feature importance analysis). See `code/README.md` for requirements and usage. ## Status This is a **partial, work-in-progress release** accompanying a manuscript currently under revision. The complete dataset (43,100 packets) will be added in a subsequent release, along with the per-node raw files needed for the Packet Delivery Rate calculation (currently pending correction — see `matlab/README.md`). For access to the complete dataset in the meantime, contact the corresponding author. ## License This repository is released under the Creative Commons Attribution 4.0 International License (CC-BY 4.0). See `LICENSE` for t …