Journal for Technology and Science
Vol. 1 No. 1 (2024): Journal for Technology and Science

AI-DRIVEN RANSOMWARE DETECTION AND PREVENTION FRAMEWORK FOR ENTERPRISE NETWORKS USING BEHAVIORAL AND NETWORK TELEMETRY

Mahbub Hasan (BSc in Information Technology, University of Information Technology and Sciences (UITS), Dhaka, Bangladesh)
Mohammad Masudur Rahman (Department of Marine Science and Oceanography, Florida Atlantic University, USA)
Mst. Murshida Khatun (B.Sc in Textile Engineering, City University, Dhaka, Bangladesh)
Muhammad Farhan Nafees (Bachelors in Business Administration, Shah Abdul Latif University, Pakistan)
Muhammad Rahim (Bachelor in English Language and Literature, International Islamic University, Pakistan)



Article Info

Publish Date
30 Dec 2023

Abstract

Objective: To address this challenge, this research presents an AI-Driven Ransomware Detection and Prevention Framework for Enterprise Networks Using Behavioral and Network Telemetry, a framework that supports the early detection and prevention of ransomware attacks by leveraging intelligent analysis of network behavior and traffic characteristics. The proposed framework aims to combine machine learning with behavioral analytics to detect malicious activities by analyzing network telemetry attributes, traffic patterns, and communication behaviors in enterprise contexts. Method: The UNSW-NB15 benchmark dataset is used to train and test the proposed framework due to the availability of a complete set of enterprise network traffic with both normal and malicious traffic having forty-nine engineered behavioral and network telemetry features. While the dataset is not 100% ransomware-specific, it is suitable for representing malicious network behaviour, which can be used to develop and test AI-based models of cyber threats. The framework consists of data preprocessing, feature engineering, model development using supervised machine learning, anomaly detection, and automated prevention mechanisms, to help boost detection efficiency while minimizing false positive rates. The performance is evaluated based on well-known cyber security metrics such as Accuracy, Precision, Recall, F1-Score, Receiver Operating Characteristic (ROC), Area under the Curve (AUC), and Confusion Matrix analysis. Results: The planned policy will boost enterprise cyber security by facilitating proactive detection of suspicious network activity that can prevent ransomware attacks from becoming a threat to organizational assets. The results should help to prove the value of artificial intelligence in today's cyber security landscapes and offer real support for enterprise-scale adoption of AI ransomware mitigation techniques. Novelty: The research provides a scalable, intelligent, data-driven security framework capable of responding to threats quickly, augmenting incident response capabilities and bolstering the resiliency of enterprise networks to the changing nature of ransomware attacks.

Copyrights © 2024






Journal Info

Abbrev

IPTEKS

Publisher

Subject

Aerospace Engineering Automotive Engineering Chemical Engineering, Chemistry & Bioengineering Civil Engineering, Building, Construction & Architecture Electrical & Electronics Engineering

Description

The Journal for Technology and Science published by Antis Publisher eISSN 3047-4337 is a scholarly journal that focuses on original research articles in natural science and technology relevant to industries and communities in developing countries. Released annually in March, August, and November, it ...