A Distributed Intrusion Detection System using Machine Learning for IoT based on ToN-IoT Dataset

The internet of things (IoT) is a collection common physical which can communicate and synthesize data utilizing network infrastructure by connecting to the internet. IoT networks are increasingly vulnerable security breaches as their popularity grows. Cyber attacks among most popular severe dangers security. Many academics interested in enhancing systems. Machine learning (ML) approaches were employed function intrusion detection systems (IDSs) provide better capabilities. This work proposed novel distributed system based on machine ML detect mitigate malicious occurrences. Furthermore, NSL-KDD or KDD-CUP99 datasets used great majority current studies. These not updated with new attacks. As consequence, ToN-IoT dataset was for training testing. It created from large-scale, diverse network. reflects each layer system, such cloud, fog, edge layer. Various methods tested specific partition dataset. model first suggested collected same all layers. Chi2 technique pick features reduced number 20. Another feature selection tool windows correlation matrix, extract relevant whole To balance classes, SMOTE method used. paper tests numerous both binary multi-class classification problems. According findings, XGBoost approach superior other algorithms node model.