NIDD-enabled lightweight intrusion detection for effective DDoS mitigation in 5G and beyond

Area of Science:

• Cybersecurity
• Telecommunications Engineering
• Machine Learning

Background:

• 5G technology promises high-speed connectivity but introduces new security challenges, particularly for Non-IP Data Delivery (NIDD).
• Robust anomaly detection is critical for safeguarding Internet of Things (IoT) and other networks against intrusions in 5G environments.
• Increasing reliance on connected technologies necessitates intelligent and efficient methods for network availability, secrecy, and integrity.

Purpose of the Study:

• To propose a novel, lightweight intrusion detection model for 5G and beyond networks.
• To address security threats such as botnet infiltration and Distributed Denial of Service (DDoS) attacks in 5G NIDD scenarios.
• To evaluate the model's effectiveness using the 5GNIDD dataset and various machine learning classifiers.

Main Methods:

• Developed a preprocessing model incorporating Gini Importance for feature selection.
• Employed state-of-the-art classifiers: AdaBoost, Easy Ensemble, GRU, 1D-CNN, LSTM, and hybrid CNN-LSTM.
• Conducted experiments using the 5GNIDD dataset, including case studies to analyze the impact of feature selection on precision.

Main Results:

• The proposed model achieved high accuracy, reaching 99.64% with 1D-CNN and hybrid CNN-LSTM classifiers.
• A precision of 0.9830 was obtained, demonstrating the model's effectiveness in identifying network anomalies.
• Experiments highlighted the effect of the curse of dimensionality on detection precision when using a reduced set of features.

Conclusions:

• The novel lightweight intrusion detection model effectively enhances security for 5G networks, particularly in NIDD scenarios.
• The model demonstrates superior performance in detecting various network incursions, ensuring data integrity and network availability.
• The findings underscore the importance of advanced anomaly detection techniques for future wireless communication systems.