Title: Implementation and experiment of the research platform in Dhaka, Bangladesh for “Real-time Malicious TLS Traffic Detection using Machine Learning Classifiers”

Currently, the use of encrypted traffic with TLS has been widely adopted. The encrypted traffic with TLS provides secure communication through the internet. However, cybercriminals also conduct their malicious activity through encrypted networks. Identifying malicious activity on an encrypted network traditionally requires large computing resources because we have to decrypt the traffic to view the traffic content, evaluate the threat, and re-encrypt it to send it back to the destination server. In addition to the issues of computing resources, decrypting traffic compromises encryption integrity and user’s privacy. This research focuses on detecting malicious internet traffic in encrypted networks using passive inspection without decryption. We will develop an application that can identify various types of TLS-based cyber-attacks on servers in the encrypted network using machine learning classifiers. Our proposed system will tap data in the edge route of campus and implement machine learning to classify the malicious traffic and then directly filter the identified malicious traffic.

Team Members

• Dr. Hossen Asiful Mustafa
• Dr. Md. Jarez Miah
• Samin Rahman Khan
• Nasir Ahmed Bhuiyan
• Md. Habibur Rahman
• Md. Mustaqim Abrar
• Shakil Ahammad

Partners

• Hasanuddin University (UNHAS)
• Universitas Brawijaya (UB)
• Universitas Syiah Kuala (USK)
• Universiti Sains Malaysia (USM)