Functional prototype for intrusion detection system oriented to intelligent iot models

Abstract

The importance of IoT (internet of things) systems, that allow things to be connected to the Internet and increase their functionalities, is becoming increasingly more relevant. The number of connected devices is growing exponentially. The special features of these devices, and the protocols used in IoT systems, make them more vulnerable to intrusion attacks. New needs arise in terms of network security. To improve the security of an IoT system without affecting the performance of the systems, an IDS (Intrusion Detection Systems) is proposed to detect anomalies in the IoT environment. In order to do so, machine learning techniques as well as the dataset used and the classification method must be taken into account. Our research focuses on the development of an IDS prototype that takes the network frames of an IoT environment using the MQTT protocol, a dataset with a compilation of attacks in a system that uses the protocol, and tests a classification model in a real time environment.