Deep Learning Model for Intrusion Detection in SDN Networks
V. Jyothsna, E. Sandhya, Thammisetty Swetha, P. Lokesh Kumar Reddy, B. Jyothsna, P. Bhasha
Abstract
Network security has grown in importance as computer network technology has progressed. Today networking technologies are vulnerable to a variety of security risks, and intrusion detection has emerged as a key tool for identifying network attacks. The architecture of Software-Defined Networks (SDN) is utilized to offer network monitoring and function observation. In order to protect computer networks and systems, IDS serves a crucial role. In order to keep the SDN’s security at a high level, an IDS is typically created to monitor the SDN’s normal traffic. A SDN network’s different security threats can be detected and managed using a number of strategies. For the purpose to identify the attacks, we combine LSTM with a deep convolution neural network (Deep CNN) driven by Grey Wolf Optimization (GWO). Effective intrusion detection requires a CNN feature extraction procedure, deep LSTM is used to detect network intrusion, and a deep CNN-LSTM is trained with a specially devised optimization strategy to improve detection performance.