Traffic Feature Selection and Distributed Denial of Service Attack Detection in Software-Defined Networks Based on...

Area of Science:

• Cybersecurity
• Network Engineering
• Machine Learning

Background:

• 5G technology's increased speed and density exacerbate network security challenges, particularly Distributed Denial of Service (DDoS) attacks in Software-Defined Network (SDN) environments.
• Network complexity in 5G can introduce noise into Intrusion Detection Systems (IDS), reducing model generalization and detection performance.

Purpose of the Study:

• To enhance the performance of Intrusion Detection Systems (IDS) in 5G networks, focusing on improving detection speed and accuracy.
• To propose and validate an innovative feature selection (FS) method for filtering network traffic data.

Main Methods:

• Developed an innovative feature selection (FS) method to identify representative and distinguishing features from network traffic data.
• Evaluated the FS method's efficacy using four common machine learning (ML) models on the InSDN, CICIDS2017, and CICIDS2018 datasets.
• Conducted real-time DDoS attack detection experiments on a simulation platform.

Main Results:

• The proposed FS technique significantly reduces detection time while maintaining or improving DDoS detection accuracy.
• The method demonstrates suitability for 5G network requirements, ensuring high speed and reliability for IDS.
• Improved robustness and detection efficiency of the IDS were observed.

Conclusions:

• The novel FS method effectively addresses the challenges of DDoS detection in complex 5G networks.
• The approach enhances IDS performance, making it suitable for high-speed, high-reliability 5G environments.
• This feature selection strategy offers a practical solution for improving cybersecurity in next-generation networks.