Saved in:
Bibliographic Details
Main Authors: Ahmet Faruk GÖRMÜŞ, Muhammed Erdem Hatayoğlu, Eltaj Pirverdiyev, Gökçe Karacayılmaz, Serkan Gönen, Deniz Dahman
Format: Recurso digital
Language:English
Published: Zenodo 2025
Online Access:https://doi.org/10.5281/zenodo.17530769
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866902260594769920
author Ahmet Faruk GÖRMÜŞ
Muhammed Erdem Hatayoğlu
Eltaj Pirverdiyev
Gökçe Karacayılmaz
Serkan Gönen
Deniz Dahman
author_facet Ahmet Faruk GÖRMÜŞ
Muhammed Erdem Hatayoğlu
Eltaj Pirverdiyev
Gökçe Karacayılmaz
Serkan Gönen
Deniz Dahman
contents Uncrewed aerial vehicles (UAVs) are effectively utilized in a wide range of applications, including reconnaissance, surveillance, logistics operations, and agricultural monitoring. However, their firm reliance on GNSS-based navigation systems presents significant cybersecurity risks. This study aims to develop an artificial intelligence-based model capable of detecting GPS jamming and spoofing attacks simulated in a controlled test environment. The attack simulations were successfully executed using a software-defined radio (SDR) device named HackRF One, and GNSS data were collected through a NEO-M8N GPS module integrated into a Pixhawk flight controller. The data collection process was conducted on an Ubuntu-based system using custom scripts, and the resulting dataset was analyzed using seven different machine learning algorithms. Among these, the XGBoost algorithm demonstrated the best performance, achieving an accuracy of 96.4%, a recall of 96.4%, a precision of 96.5%, and an F1-score of 96.4%. With a training time of approximately 1.03 seconds and a test time of 0.01 seconds, the model proved suitable for real-time applications. The developed AI-based system was integrated into the flight controller via Raspberry Pi and successfully tested under real flight scenarios, allowing instant detection and response during an attack. In conclusion, the proposed approach offers a cost-effective, reliable, and field-compatible solution that enhances the cybersecurity of both civilian and military UAV operations.
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_17530769
institution Zenodo
language eng
publishDate 2025
publisher Zenodo
record_format zenodo
spellingShingle Real-Time Detection of GPS Dimming and Fraud Attacks in Unmanned Aerial Vehicles with Artificial İntelligence
Ahmet Faruk GÖRMÜŞ
Muhammed Erdem Hatayoğlu
Eltaj Pirverdiyev
Gökçe Karacayılmaz
Serkan Gönen
Deniz Dahman
Uncrewed aerial vehicles (UAVs) are effectively utilized in a wide range of applications, including reconnaissance, surveillance, logistics operations, and agricultural monitoring. However, their firm reliance on GNSS-based navigation systems presents significant cybersecurity risks. This study aims to develop an artificial intelligence-based model capable of detecting GPS jamming and spoofing attacks simulated in a controlled test environment. The attack simulations were successfully executed using a software-defined radio (SDR) device named HackRF One, and GNSS data were collected through a NEO-M8N GPS module integrated into a Pixhawk flight controller. The data collection process was conducted on an Ubuntu-based system using custom scripts, and the resulting dataset was analyzed using seven different machine learning algorithms. Among these, the XGBoost algorithm demonstrated the best performance, achieving an accuracy of 96.4%, a recall of 96.4%, a precision of 96.5%, and an F1-score of 96.4%. With a training time of approximately 1.03 seconds and a test time of 0.01 seconds, the model proved suitable for real-time applications. The developed AI-based system was integrated into the flight controller via Raspberry Pi and successfully tested under real flight scenarios, allowing instant detection and response during an attack. In conclusion, the proposed approach offers a cost-effective, reliable, and field-compatible solution that enhances the cybersecurity of both civilian and military UAV operations.
title Real-Time Detection of GPS Dimming and Fraud Attacks in Unmanned Aerial Vehicles with Artificial İntelligence
url https://doi.org/10.5281/zenodo.17530769