Explorando abordagens de segurança para detecção de intrusão utilizando aprendizado de máquina em internet das coisas

Abstract

The present work aims to analyze the use of state-of-the-art methods that involve Machine Learning (ML) techniques for defending against intrusion attacks in Internet of Things (IoT) networks. The study seeks to understand how research addressing defense mechanisms against intrusion attacks has evolved, selecting studies that focus on the use of ML as a protection approach, as well as identifying the main types of attacks and their motivations aimed at compromising IoT networks. Based on the selected articles, it was observed how defense applications employing ML techniques are applied, through the use of four data mining methods: analysis of the frequency of specific terms; the Apriori method for item frequency analysis; the clustering method for identifying groups with similar characteristics; and time series analysis to observe the behavior of the analyzed articles over the defined research period. In this way, it was possible to formulate three research questions addressing the challenges identified in this domain and how they may be mitigated through the use of ML in defending against intrusion attacks in IoT networks.