Approaches on intrusion detection integrating with deep learning in sdn and iot

Abstract

Analyzing network traffic data is essential for the effective communication of information through communication systems. Due to the frequent occurrence of network intrusions, ensuring robust network security is more important. This involves protecting data from unauthorized access by identifying and preventing potential attacks. An Intrusion Detection System (IDS) is a key technology in this effort, as it continuously monitors network traffic to detect any potential threats. A lot of research has been conducted on IDS to distinguish between legitimate and unauthorized users. However, existing methods face challenges when dealing with binary classification and multi-class scenarios, as they tend to focus on the features selection while excluding categorical ones. To address these shortcomings, this thesis addresses the critical need for robust IDS within Software-Defined Networking (SDN) and Internet of Things (IoT) environments. In Chapter 1, the fundamental architecture of SDN and IoT is discussed, emphasizing the need for effective IDS to secure these networks. This chapter also provides an overview of the benchmark dataset used in the research, highlighting its characteristics and relevance. newline