Minority Resampling Boosted Unsupervised Learning With Hyperdimensional Computing for Threat Detection at the Edge of Internet of Things

The Internet of Things (IoT) has rapidly transformed digital environments across a multitude domains with increased connectivity and pervasive virtualization. distributed computing paradigm Edge Computing been postulated to overcome the concerns response time, bandwidth, energy consumption, cybersecurity. In comparison other concerns, limited studies have focused on cybersecurity, mainly due inherent complexity threat detection at Edge. However, widespread adoption IoT applications in economic, social, political contexts is stringent indication significant impact from cyber-attacks. This paper aims address this challenge by presenting an effective efficient machine learning approach for IoT. novel contributions are, new Enhanced Geometric Synthetic Minority Oversampling Technique (EG-SMOTE) algorithm resolve imbalanced distribution data streams Edge, extension Growing Self Organizing Map (GSOM) based Hyperdimensional unlabeled streams. proposed EG-SMOTE + GSOM tested using four open access datasets; three benchmark, KDD99 (F-Score = 0.9360), NSL-KDD 0.9647), CICIDS2017 0.9999), one industry-focused botnet traffic dataset, BoT-IoT 0.9445). outperformed SMOTE G-SMOTE approaches vast number experiments that are tried different classifiers. results these confirm novelty, efficiency effectiveness cybersecurity