A Cloud Based Honeycloud System For Malicious Detection Using Machine Learning Techniques

Abstract

With the massive expansion of IoT botnet DDoS attacks in recent years, IoT protection has now become one of the most concerned topics in the area of network security. In this paper, we propose a honeypot-based method that uses machine-learning techniques for malware detection in the IoT system. The IoT honeypot developed data is used as a dataset for the practical and dynamic training of a machine learning model. The honeypots are developed and placed in a cloud network that allows us to gather the unknown and known incidents in cloud computing. In our cloud network, we suggest a HoneyCloud system that highlights catching any attack or suspect action on protocols like Secure Shell (SSH) protocol, File Transfer Protocol (FTP), etc. These methods can be employed to define the distinction between Malicious and benign traffic. Also, we implemented various machine learning models in this paper and compared them with the parameters of True positive (TP) and false positive (FP) rates; this comparative analysis on which one of these machine learning-based classification algorithms would give us a low false positive rate.