Deep learning-based Internet of Things intrusion detection

Abstract

A number of models use deep learning to find new ways to infiltrate more secure networks and identify Internet of Things (IoT) attacks. The nature of IoT data and its growing applications, which make attacks more common, has increased the need for the development of an intrusion detection system to quickly identify and categorize attacks. Malicious assaults are always developing and changing. In this research, we investigate how to distinguish between legitimate and malicious behavior while analyzing network data for new threats in order to identify abnormalities and intrusions. This study analyzes earlier work and evaluates the efficacy of previous studies utilizing two fresh types of current traffic data (For example, Bot-IoT and CSE-CIC-IDS2018 datasets). We provide accuracy tests for intrusion detection in various systems to assess performance.