Related Experiment Video
Updated: May 15, 2025
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
436
A secure and efficient deep learning-based intrusion detection framework for the internet of vehicles.
Hasim Khan1, Ghanshyam G Tejani2,3, Rayed AlGhamdi4
1Department of Mathematics, College of Science, Jazan University, Kingdom of Saudi Arabia, P.O. Box 114, 45142, Jazan, Saudi Arabia.
Scientific Reports
|April 10, 2025
Summary
This study introduces a novel intrusion detection system (IDS) for Internet of Vehicles (IoV) security, enhancing data privacy and real-time threat detection. The system achieves high precision, ensuring secure and reliable IoV network operations.
Area of Science:
- Computer Science
- Cybersecurity
- Network Engineering
Background:
- Internet of Vehicles (IoV) networks face significant security challenges due to their complex and dynamic nature, necessitating robust intrusion detection systems (IDS).
- Existing security measures often struggle to balance real-time performance, data privacy, and effective threat identification in large-scale IoV environments.
Purpose of the Study:
- To develop an advanced, secure, and highly precise intrusion detection system tailored for Internet of Vehicles (IoV) environments.
- To integrate innovative cryptographic techniques and deep learning models for enhanced IoV security and privacy-preserving data processing.
Main Methods:
- Implementation of AES-256 encryption, Secure Multi-Party Computation (SMPC), and Homomorphic Encryption (HE) for secure data handling.
- Utilizing Z-score normalization and median imputation for data preprocessing, followed by Vision Transformer (ViT), wavelet transforms, and GAT for feature extraction.
- Introduction of a novel Crayfish-Mother secure Optimization (CMSO) method for optimized feature selection and a DAGSNet architecture integrating DenseNet, GoogleNet, AlexNet, and SqueezeNet for improved detection and classification.
Main Results:
- The proposed IDS demonstrates high effectiveness with maximum precision rates of 0.991 and 0.984 on two datasets.
- Achieved minimal encryption and decryption times of 0.02s and 0.82s, respectively, ensuring real-time security.
- The holistic approach significantly enhances the dependability and effectiveness of intrusion detection in IoV networks.
Conclusions:
- The developed IDS provides a highly secure, effective, and precise solution for IoV environments, addressing critical security and privacy concerns.
- The integration of advanced encryption, privacy-preserving computation, and deep learning models offers a robust framework for future IoV security research.
- This work establishes a new benchmark for intrusion detection systems in the context of rapidly evolving Internet of Vehicles technology.
Keywords:
Crayfish-Mother swarm optimizerDAGSNetDeep learningHybrid optimizationInternet of vehiclesIntrusion detection
