Using Machine Learning for malware traffic prediction in IoT networks

IoT devices have become the mainstream technology in many industries. Typically, these devices have permanent network connection which makes them vulnerable to outside attacks. In many cases security mechanisms cannot be built-in into each IoT device due to limited computational power available. Instead, the existing Intrusion Detection / Prevention Systems can use machine learning to protect networks that carry IoT traffic. This paper shows how machine learning can be used to detect malicious traffic in IoT networks. The study used IoT-23 dataset and the accuracy recorded ranges from 98.9%–100% depending on the sub-dataset.