Intrusion of Malware (DDoS) detection in IoT devices using Machine Learning on Cyberspace

Abstract

Nowadays, the number of interconnected devices (IoT) is increasing dramatically. This expansion poses new security problems for network operators, IoT service providers, and users. Security measures implemented on IoT devices are getting complex due to their heterogeneity and constraints. Attackers have utilized IoT devices to execute massive attacks like DDoS, Zero-Day-Exploitation, Ransomware, etc. The most significant measure to safeguard services from insecure IoT devices is to increase security consciousness in the core network. On the other hand, this thesis suggests a machine learning DDoS detection and diminution technique. The proposed approach was assessed by applying five supervised machine learning classification methods. The evaluation findings reveal that k-NN and Random Forest algorithms outperform ANN, SVM, and Naïve Bayes algorithms. Consequently, the findings of this study can assist academics in further research on malware detection systems for IoT devices.