Towards a rooted subgraph classifier for IoT botnet detection

Abstract

The Internet of Things (IoT) devices provide various benefits for our modern life. However, in recent years, commercial-off-the-shelf devices such as IP-Camera, Router, Smart-TV, etc. are being targeted more and more by IoT Botnet. Therefore, the detection of IoT botnet malware is essential. Recently, some of the studies have used machine learning and deep learning for the automatic detection of malware. However, machine learning and deep learning also have their own advantages and disadvantages. Therefore, in this paper, we have proposed a method that combine deep learning and machine learning to generate a novel feature-based PSI-Rooted sub-graph for detecting cross-architecture IoT botnet malware. This feature is robust enough for various common machine learning classifiers that achieved an accuracy of about 97% and F-score about 98%.