An Innovative Approach for Efficient Detection and Classification of Malware in 5G-IoT Healthcare Systems
L. Bharathi, L. Bhagyalakshmi, S Nagakishore Bhavanam, M. Sindhuja, K. Sravan Abhilash, Amit Lathigara · 2025
Utilized unique Convolutional Neural Network (CNN) models to provide a strong strategy for combating malware in 5G-IoT healthcare applications. Protecting the 5G-IoT healthcare facilities, the suggested module can identify and categorize various forms of malware with high accuracy. In order to help cybersecurity professionals, comprehend the threat environment and develop suitable defences, the categorization CNN model sorts the discovered malware into certain families. Additionally, the precision, losses, and additional metrics of several CNN models are compared. In order to analyse the wellness of patients in real time, innovative applications for healthcare need to monitor them remotely. Nevertheless, there are also a number of issues that have been discovered with the use of 5G-IoT for health care use cases. These include a lack of cost-effective assets, an are lacking of a standard structure for handling programs, and the primary worry of malicious information or the security of private medical information. Another issue is the waste of communication bandwidth caused by a lack of network bandwidth improvement. This issue is addressed by proposing a four-module improved system structure. The suggested CNN approach outperforms Deep Q-Learning (Deep-Q) , MLP, BPNN, and SVM in terms of total accuracy, which stands at 98.30%. As a result, CNN is clearly the best option for malware categorization.