Cloud-based Detection of Malware and Software Privacy Threats in Internet of Things using Deep Learning Approach

Abstract

The term “cyber-physical system” (CPS) refers to integrating computational and communication capabilities with physical processes. Because of patient data's regulatory and ethical implications, cybersecurity has emerged as a critical issue in the healthcare industry. Because of the sensitive nature of patient information, the layout of CPS models for any large databases requires extra precautions. Protecting user privacy and fending off attacks like spoofing, DoS, jamming, and eavesdropping are essential for cloud storage, which integrates multiple databases to deliver cutting-edge, intelligent services. This manuscript proposes a hybrid deep-learning method for scanning the entire IoT network for malware and pirated software. It is suggested that a Deep learning deep neural network be used to detect source code plagiarism in pirated software. Source code plagiarism is filtered through tokenisation and weighted feature methods, magnifying each token's significance. Next, we use a deep learning method to check for copied code. The data comes from Google's Code Jam (GCJ), and it was gathered with the intention of studying software theft. In addition, malicious infections in an IoT network can be detected by means of colour image visualisation using a deep convolutional neural network. Malware samples from the Maling dataset are used in the experiments. The experimental results show that the proposed methodology outperforms state-of-the-art methods in terms of classification performance when gauging the severity of cybersecurity threats in the Internet of Things (IoT).