Classifying Tor Traffic Encrypted Payload Using Machine Learning

Project Code :TCMAPY1238

Objective

This project aims to develop and evaluate a machine learning-based framework for classifying Tor traffic encrypted payloads, with the goal of enhancing cybersecurity. By leveraging a dataset characterized by attributes like Source and Destination Ports, Protocol, Flow Duration, and various Inter-Arrival Times, we implement and assess the performance of Decision Tree, Logistic Regression, and XGBoost classifiers. The project focuses on accurately identifying the nature of network traffic to distinguish between benign and malicious activities, thereby promoting a more secure and efficiently monitored network environment.

Abstract

The rapid evolution of internet technologies has necessitated advanced methodologies for monitoring and classifying encrypted network traffic. This study introduces a robust framework utilizing Machine Learning (ML) to classify Tor traffic encrypted payloads, an essential step for enhancing cybersecurity measures. Utilizing a dataset featuring columns such as Source Port, Destination Port, Protocol, Flow Duration, various Inter-Arrival Times (IAT), and others, we apply three distinct ML models: Decision Tree, Logistic Regression, and XGBoost. Our objective is to accurately predict the nature of traffic ('label' as the target column), thereby distinguishing between benign and potentially malicious activities. The effectiveness of each model is evaluated based on their predictive accuracy and computational efficiency, offering insights into the optimal approaches for real-time encrypted traffic analysis. This research contributes to the development of more secure network environments by leveraging advanced data analytics in the realm of cybersecurity.

Keywords: Decision Tree Classifier, Logistic Regression and XGBoost Classifier.

NOTE: Without the concern of our team, please don't submit to the college. This Abstract varies based on student requirements.