A Deep Learning Framework for Robust Malware Detection in Wireless Communication Networks
Walid Abushiba, Princy Johnson, Brahim Benbakhti, Kashif Jamshid, Muhammad Irfan, Anaum Ihsan
Abstract
Wireless communication networks are increasingly vulnerable to advanced malware attacks, posing significant risks to network security and user privacy. Traditional detection methods struggle with sophisticated malware, necessitating approaches that are more robust. This study introduces a deep learning-based detection system that leverages Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Graph Convolutional Networks (GCNs) to analyze network traffic and detect malicious activity. Results demonstrate that these models outperform traditional machine learning in accuracy, precision, and recall, offering improved resilience against emerging malware threats. The findings underscore deep learning's potential to enhance wireless network security and pave the way for future innovations in malware detection. Our proposed models achieved up to 96.3% accuracy and 98.2% precision, outperforming traditional machine-learning approaches by a significant margin.