DEEP LEARNING MODELS FOR ADVANCED INTRUSION DETECTION IN NEXT-GENERATION NETWORKS

The rapid evolution of next-generation networks like Software Defined Networks (SDN), Internet of Things (IoT), and 5G infrastructures has made cybersecurity issues extremely complex. The traditional intrusion detection system (IDS) mostly depends upon signature-based intrusion detection techniques that fail to detect sophisticated and unknown cyber attacks. Therefore, the integration of deep learning techniques with intrusion detection has become a promising solution to improve network security. In this paper, a deep learning-based intrusion detection framework has been proposed to detect complex and unknown attacks in next-generation networks. The proposed framework uses a hybrid deep learning architecture that combines Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks to detect unknown attacks in next-generation networks. The proposed framework has been evaluated using benchmark datasets like NSL-KDD and UNSW-NB15 datasets that contain different categories of network attacks like DoS, Probe, R2L, and U2R attacks.