Multi-Layer Perceptron Artificial Neural Network Based IoT Botnet Traffic Classification
share
Internet of Things (IoT) is becoming an integral part of our homes today. Internet-connected devices, such as smart speakers, smart bulbs, and security cameras are improving our convenience and security. With the growth in smart environments, there is an increasing concern over the security and privacy issues related to IoT devices. The issue of the IoT security has received considerable attention due to (1) the intrinsic technological constraints of IoT devices (computing and storage limitations) and (2) its prevalence in people’s life’s, in close proximity. IoT devices can be easily compromised (much easier than PCs and/or smart phones) and can be utilized for generating botnet attacks. In this paper, we propose an Artificial Intelligence (AI) based solution for malicious traffic detection. We explore the accuracy of Multi-Layer Perceptron (MLP) Artificial Neural Network (ANN) learning algorithm in detecting botnet traffic from IoT devices infected by two major IoT botnets, namely, Mirai and Bashlite (also known as Gafgyt). After tuning and optimization, the MLP-ANN algorithm achieved an accuracy rate of 100% in the testing phase of IoT botnet traffic classification.
READ FULL TEXT