Cyberattack Detector for Real-Time IoT Devices and Networks

The Internet of Things is one of the great technological revolutions of recent times. With interconnected devices and, in some cases, with sensitive information, it is essential to have a defense mechanism capable of guaranteeing the security of devices and networks. In this article, we propose a binary classifier based on simple machine learning techniques with a low computational impact that allows the detection of cyberattacks. These techniques will facilitate implementation on devices with low resources, allowing constant status monitoring. This classifier is validated with a dataset of cyberattacks from a real-time IoT infrastructure through a 10K-Fold cross-validation and a subsequent validation stage. The results are promising, indicating a robust and reliable detection of cyberattacks.