ROBUST AND LIGHTWEIGHT INTRUSION DETECTION MODEL FOR ATTACK CLASSIFICATION IN 5G ENVIRONMENTS

Abstract

The advent of 5G networks has introduced unprecedented challenges in security due to their complex architecture, increased connectivity, and reliance on software-defined components. Traditional IDS struggle to meet the demands of real-time threat detection with minimal latency. This paper proposes a robust and lightweight IDS that leverages extensive data analytical techniques along with machine learning (ML), deep learning (DL), and hybrid approaches for effective attack classification in 5G networks. The proposed system integrates feature selection, ensemble learning, and optimization strategies to enhance detection accuracy while minimizing computational overhead. Experimental validation on benchmark datasets demonstrates the model's effectiveness in identifying and mitigating various attack vectors, including DDoS, MITM, and malware injections, with high precision and recall.