Drone detection with radio frequency signals and deep learning models is an open-source paper by Xuanze Dai.
The widespread use of drones raises significant security, environmental, privacy, and ethical concerns, making effective drone detection crucial. Various methods, such as wireless signal detection, photoelectric detection, radar detection, and sound detection, are currently employed to detect drones. However, these methods often lack the precision needed for accurate drone identification. To address this challenge, more robust detection techniques are required. Additionally, different types of drones and application scenarios necessitate tailored detection and identification approaches.
This study compared the performance of different machine learning, deep learning, and multi-task models on an open radio frequency (RF) signal dataset across 2-class, 4-class, and 10-class problems. By integrating RF methods with Convolutional Neural Networks (CNNs), the author developed a multi-task model for drone detection and classification. The experimental results demonstrate that the XGBoost model achieved exceptional accuracy, with 99.96% for the 2-class problem, 92.31% for the 4-class problem, and 74.81% for the 10-class problem, making it the best-performing model for drone detection and classification on this groundbreaking dataset.
Publication Date– March 2024
Drone detection with radio frequency signals and deep learning models contains the following major sections:
- Introduction
- Related Work
- Dataset Description
- Methods
- Results
- Conclusion
Post Image Credit- envatoelements by lzf
