IOT Security Against Network Anomalies through Ensemble of Classifiers Approach

Abstract

The use of IoT networks to monitor critical environments of all types where the volume of data transferred has greatly expanded in recent years due to a large rise in all forms of data. Since so many devices are connected to the Internet of Things (IoT), network and device security is of paramount importance. Network dynamics and complexity are still the biggest challenges to detecting IOT attacks. The dynamic nature of the network makes it challenging to categorise them using a single classifier. To identify the abnormalities, we therefore suggested an ensemble classifier in this study. The proposed ensemble classifier combines the independent classifiers ELM, Nave Byes (NB), and the k-nearest neighbour (KNN) in bagging and boosting configurations. The proposed technique is evaluated and compared using the MQTTset, a dataset focused on the MQTT protocol, which is frequently utilised in IoT networks. The analysis demonstrates that the proposed classifier outperforms the baseline classifiers in terms of classification accuracy, precision, recall, and F-score.