Deep Complex Gated Recurrent Networks-Based IoT Network Intrusion Detection Systems.

Area of Science:

• Cybersecurity
• Network Security
• Artificial Intelligence

Background:

• The rapid expansion of the Internet of Things (IoT) necessitates advanced security solutions.
• Traditional Intrusion Detection Systems (IDS) struggle with the unique challenges of IoT environments, including device diversity and real-time detection needs.

Purpose of the Study:

• To propose DCGR_IoT, a novel deep neural learning-based IDS designed for bidirectional IoT communication networks.
• To enhance anomaly detection capabilities within IoT environments.

Main Methods:

• Utilizing Convolutional Neural Networks (CNN) for spatial feature extraction and data filtering.
• Employing Complex Gated Recurrent Networks (CGRNs) for temporal feature extraction and multidimensional feature subset construction.
• Leveraging CGRNs to create detailed spatial representations of network traffic for critical feature extraction.

Main Results:

• DCGR_IoT demonstrated high effectiveness on benchmark datasets (UNSW-NB15, KDDCup99, IoT-23).
• Achieved a superior detection accuracy rate of 99.2% against sophisticated cyber-attacks.
• Validated the system's capability for efficient and accurate intrusion detection in IoT networks.

Conclusions:

• DCGR_IoT presents a robust and effective solution for safeguarding IoT networks.
• The proposed system addresses the limitations of conventional IDS in dynamic IoT environments.
• Highlights the potential of deep learning models for advanced IoT security.