Enhancing Security in 5G Edge Networks: Predicting Real-Time Zero Trust Attacks Using Machine Learning in SDN...

Area of Science:

• Cybersecurity
• Machine Learning
• Network Security

Background:

• The internet faces increasing cyber threats, including sophisticated Distributed Denial of Service (DDoS) attacks.
• Traditional security systems struggle to detect advanced DoS and DDoS attacks effectively.
• Machine learning (ML) shows promise for enhanced attack detection, yet real-time capabilities remain a challenge.

Purpose of the Study:

• To develop and evaluate a real-time system for detecting Distributed Denial of Service (DDoS) attacks.
• To address the limitations of current security solutions in identifying complex network intrusions.
• To leverage machine learning for accurate and rapid identification of cyber threats.

Main Methods:

• A simulated network environment was created using Mininet and the POX Controller.
• The CICDDoS2019 dataset was utilized for attack identification and classification.
• Pre-trained machine learning models analyzed network traffic in real-time within a virtual software-defined network (SDN).

Main Results:

• The proposed methodology achieved a 99% accuracy rate in detecting DDoS attacks.
• The system demonstrated a rapid detection time, identifying attacks within 1 second.
• The model successfully classified and identified various DDoS attack types in the simulated environment.

Conclusions:

• The developed machine learning-based system offers a highly accurate and efficient solution for real-time DDoS attack detection.
• The integration of ML with SDN provides a robust framework for enhancing network security against advanced threats.
• This research contributes to bridging the gap in real-time detection of complex cyberattacks.