DroneRF dataset: A dataset of drones for RF-based detection, classification and identification.

Area of Science:

• Computer Science
• Electrical Engineering
• Signal Processing

Background:

• The proliferation of drone technology necessitates advanced methods for monitoring and security.
• Effective data analysis is crucial for understanding complex systems and informed decision-making.
• Radio frequency (RF) signals offer a unique signature for identifying and tracking aerial devices.

Purpose of the Study:

• To introduce and describe the DroneRF dataset, a novel collection of RF data from drones.
• To provide a resource for research in drone detection, identification, and RF signal analysis.
• To facilitate the development of machine learning models for drone monitoring applications.

Main Methods:

• Collected 227 segments of RF activity from three distinct drones operating in various modes (off, on, connected, hovering, flying, video recording).
• Recorded background RF activities without drone presence for comparative analysis.
• Utilized RF receivers connected to laptops via PCIe cables to capture, process, and store data in a database.

Main Results:

• The DroneRF dataset comprises comprehensive RF recordings, including drone-specific signals and ambient RF noise.
• The dataset captures drone communications with their flight control modules.
• Provides a foundation for developing and testing RF-based drone detection algorithms.

Conclusions:

• The DroneRF dataset is a valuable resource for advancing research in drone detection and identification.
• The dataset enables the exploration of deep learning approaches for analyzing RF signals.
• Facilitates the creation of open-source databases for drone-related RF data.