Combined RF-based drone detection and classification is an open-source work by Sanjoy Basak, Sreeraj Rajendran, Sofie Pollin, and Bart Scheers.
While drones have numerous positive applications, they are also increasingly being used for illicit activities such as drug trafficking, firearm smuggling, and threatening security-sensitive locations like airports and nuclear power plants. Current drone localization and neutralization technologies assume the drone has already been detected and classified. Although significant progress in sensor technology has occurred over the past decade, a robust method for drone detection and classification has yet to be established.
This paper focuses on radio frequency (RF) based drone detection and classification by analyzing the frequency signature of the transmitted signal. The authors created a novel RF dataset using commercial drones and conducted a detailed comparison between a two-stage detection and classification framework and a combined approach. The performance of both frameworks was evaluated for scenarios involving a single signal and simultaneous multi-signal detection. Their analysis demonstrates that the You Only Look Once (YOLO) framework offers superior detection performance compared to the Goodness-of-Fit (GoF) spectrum sensing method in multi-signal scenarios while also delivering classification results on par with the Deep Residual Neural Network (DRNN) framework.
Publication Date– March 2022
Combined RF-based drone detection and classification contains the following major sections:
- Introduction
- Problem Statement
- Background
- Technical Approach
- Experiments
- Performance Analysis
- Conclusion
Post Image Credit- envatoelements by wirestock
