Related Experiment Video
Updated: Jun 6, 2026
06:19
Low-Cost Automated Flight Intercept Trap for the Temporal Sub-Sampling of Flying Insects Attracted to Artificial Light at Night
Published on: December 29, 2021
BPBiLSTM-IDS: a lightweight intrusion detection framework for cyber-physical UAV networks.
Hafiz Muhammad Attaullah1, Inam Ullah Khan1,2, Muhammad Mansoor Alam3,4
1Faculty of Computing and Informatics, Multimedia University, Cyberjaya, Malaysia.
Scientific Reports
|June 4, 2026
Summary
This study introduces an enhanced Intrusion Detection System (IDS) for Unmanned Aerial Vehicles (UAVs). The novel approach significantly improves cybersecurity by accurately detecting network attacks with high precision and recall.
Area of Science:
- Cybersecurity
- Artificial Intelligence
- Aerospace Engineering
Background:
- Unmanned Aerial Vehicles (UAVs) are vulnerable to cybersecurity threats like De-authentication Denial of Service and False Data Injection (FDI) due to wireless network reliance.
- Conventional Intrusion Detection Systems (IDS) exhibit high false alarm rates and resource inefficiency, unsuitable for dynamic UAV environments.
Purpose of the Study:
- To develop a scalable, adaptive, and lightweight Intrusion Detection System (IDS) for enhanced cybersecurity in Unmanned Aerial Vehicle (UAV) networks.
- To address the limitations of traditional IDS by improving accuracy and reducing resource consumption in UAV network traffic analysis.
Main Methods:
- Feature selection using Binary Pigeon Optimization (BP) to identify optimal features independent of computational cost.
- Intrusion detection utilizing a Bidirectional Long Short-Term Memory (Bi-LSTM) network combined with a 1D Convolutional Neural Network (1D-CNN) to capture temporal network traffic characteristics.
Main Results:
- The proposed BP + Bi-LSTM + 1D-CNN model achieved a high accuracy of 98.74% ± 0.07 on a cyber-physical UAV dataset.
- Demonstrated superior performance over traditional machine learning models (SVM, DT, RF, DNN) in terms of accuracy, precision, recall, and false positive rate.
- The model proved to be a scalable, adaptive, and lightweight solution for real-time intrusion detection in UAV networks.
Conclusions:
- The enhanced IDS architecture effectively mitigates cybersecurity threats in UAV networks.
- The combination of BP for feature selection and Bi-LSTM with 1D-CNN for detection offers a robust and efficient solution for real-time intrusion detection.
- The proposed model represents a significant advancement in securing autonomous aerial systems against sophisticated cyberattacks.
