LoRaMob: An Open Dataset for Mobile LoRaWAN

This repository contains two datasets collected from emulated LoRaWAN networks: one spanning 11 days with ADR and another covering 28 days without ADR.

The full dataset is hosted on Nextcloud due to its large size. In this repository, we provide a 3-day example dataset and scripts to process and analyze the data.

🎬 Demo Video

Click here to view the video.

🎞️ Demo Animation

---

🗂 Dataset Access and Structure

📁 What’s in GitHub?

The GitHub repository includes:

1. Example Data:

   • A 3-day subset of the dataset in CSV format.
     • Located in example_data/.
       • with_adr/: Contains 3 days of data with ADR.
       • without_adr/: Contains 3 days of data without ADR.

2. Scripts:

   • Tools for processing and analyzing the data.
     • app_data_json_to_csv.py: Converts JSON files to CSV.
     • combine_csv.py: Merges multiple CSV files into one.
     • remove_duplicates.ipynb: Removes duplicate records from the dataset.

3. Documentation:

   • This README.md explains the dataset and repository structure.
   • Guidelines for getting started with the dataset and scripts.

---

🌐 What’s on Nextcloud?

The full dataset is hosted on Nextcloud due to its large size. Download it using this link. To get the password for access, contact moheed-ali.kayani@cnam.fr.

Full Dataset Structure:

• 11 Days with ADR:
  • Infrastructure dataset:
    • Monitoring (JSON)
    • Prometheus (JSON)
  • Network dataset:
    • Application (JSON)
    • Gateway (JSON)
• 28 Days without ADR:
  • Infrastructure dataset:
    • Monitoring (JSON)
    • Prometheus (JSON)
  • Network dataset:
    • Application (JSON)
    • Gateway (JSON)

---

🚀 Getting Started

1. Explore the GitHub Repository

• Start with the 3-day example dataset in example_data/.
• Use the provided scripts to process and analyze the data.

2. Download the Full Dataset

• Access the complete dataset from Nextcloud.

3. Process the Data

• Use the scripts included in the scripts/ directory:
  • app_data_json_to_csv.py: Converts application JSON files to CSV.
    Ensure paths in the script are updated before use.
  • combine_csv.py: Merges multiple CSV files into one.
  • remove_duplicates.ipynb: Removes duplicate records caused by overlapping buffers during data collection.

4. Analyze the Data

• Explore up events or the rest of the dataset as per your analysis goals.

---

📜 Dataset Description

The dataset was generated using:

• Mobility data from bike-sharing endpoints and travel times.
• Emulation tools like elora, a LoRaWAN module for NS-3 designed for massive networks (elora GitHub Repository).
• Network and application servers implemented using ChirpStack (chirpstack-docker).

Characteristics:

• Emulated for 4,000 devices and 7 gateways.
• Covers an area of 116.877 km².
• Includes:
  • 11 days with ADR: Emulated network behavior with ADR.
  • 28 days without ADR: Emulated network behavior without ADR.

---

✨ Usage Policy

Any use of this dataset, whether for public or private purposes, must include a reference to this repository:

LoRaMob: An Open Dataset for Mobile LoRaWAN
🔗 https://github.com/cedric-cnam/LoRaMob

---

🛠 Scripts

The repository provides the following scripts for working with the dataset:

• app_data_json_to_csv.py: Converts application JSON files into CSV format.
• combine_csv.py: Merges multiple CSV files into one.
• remove_duplicates.ipynb: Removes duplicate records from the dataset.

---

🤝 Acknowledgments

This work was funded by the French ANR INTELLIGENTSIA project (grant nb: ANR-20-CE25-0011).
We thank the contributors and collaborators for their support in the data collection and processing.

---

📧 Contact

For questions or feedback, please create an issue in this repository or contact us at [moheed-ali.kayani@cnam.fr].