Metaheuristics for Machine Learning

6Ensemble Machine Learning-Based Botnet Attack Detection for IoT Applications

Suchithra M.

Department of Computing Technologies, School of Computing, SRM Institute of Science and Technology, Kattankulathur, Chennai, India

Abstract

The Internet of Things (IoT) systems are currently the subject of a rising prevalence of attacks as a result of their fast development and widespread adoption. It has been claimed that attacks using botnets account for the majority of those in IoT networks. Because the majority of IoT systems possess the memory and processing power necessary for effective security methods, there are still several security vulnerabilities in these systems. Additionally, attackers are capable of circumventing many of the present rule-based detection algorithms. Even though the machine learning (ML)-driven method can identify the variations of the various types of attacks, periodically special new types of attacks can be initiated. As a result, this paper suggests a novel ML technique termed random tree-adaptive artificial neural network (RT-ANN) in detecting botnet attacks. Initially, we collect the raw data samples and these samples are normalized. Principal component analysis (PCA) is used to retrieve key aspects of the RT-ANN. Our proposed RT-ANN method’s performance is analyzed, and comparative analysis is also provided. The experimental results demonstrate that the proposed method achieves the highest detection rate in the botnet attack detection.

Keywords: Internet ...