WITHDRAWN: A deep-RNN and meta-heuristic feature selection approach for IoT malware detection

• DOI: 10.1016/j.matpr.2021.01.207
• Publication Date: 2021-02-18T06:38:44Z
• Authors: G. Radhakrishnan; K. Srinivasan; S. Maheswaran; K. Mohanasundaram; D. Palanikkumar; Abhay Vidyarthi
• Keywords: 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology

Abstract Increasingly, Internet of Things (IoT) is used in various sectors and for various objectives. The increasing existence in huge range of technologies, and growing computing and computing capabilities build them a worth able target for attack, such as virus model to attack particular IoT systems. In this research article, we describe how deep learning techniques are having performed to identify IoT malware. In specific, our proposed method uses RNN for the study of execution process codes in ARM-based IoT frameworks. We use an IoT malware sample dataset with 271benign-ware and 282 malware to practice our model. Next, we tested the trained method by using the 104 unknown (LSTM) Long Short Term Memory samples from IoT threats. The K-fold-cross validation analysis showing the second configuration of second-layer neurons has a better accuracy of 99.08% in the detection of unknown malware. The LSTM solution provides the best probable result in a comparative description of other machine learning methods.