Analysing Cloud DDoS Attacks Using Supervised Machine Learning

##plugins.pubIds.doi.readerDisplayName## https://doi.org/10.70593/978-93-49307-78-0

Authors

Chisom Elizabeth Alozie
University of the Cumberlands, United States
https://orcid.org/0009-0000-0309-3650

Keywords:

Cloud Security, DDoS Attacks, Machine Learning, Cybersecurity, Cloud Computing, Network Security, Data Protection, Threat Analysis, Attack Detection, Security Vulnerabilities, Data Privacy, Cyber Threats

Synopsis

Cloud computing in its simplest form refers to the provision of hardware and software to deliver a service over an internet network. However, Cloud Computing has numerous issues, such as security attacks and distributed denial of service (DDoS). A DDoS attack is defined as a method of attack in which numerous computer systems are allowed to attack a target, such as a server, any resource, or website, resulting in a denial of service for the resource's intended users.

This research analysed the normal traffic and DDoS attack traffic from cloud environments using machine learning technology to detect DDoS attacks. This work’s main contribution is the extraction of dataset features and the discovery of new flow features for DDoS attack detection. To create the dataset, novel features are stored in a CSV file using the CICFlowMeter tool. Features were selected using a correlation coefficient to get better model accuracy. Machine learning algorithms were trained on the resulting cloud dataset. The existing work reviews for detection of DDoS attacks either used a cloud dataset or another network data set, or the research findings were kept confidential. The methodology used to solve this problem is the CRISP-DM methodology.

The proposed solution deployed a brand-new dataset with five machine-learning models for classification. The findings of this study help to improve knowledge of the ability of DDoS datasets to detect intrusions. Five performance metrics—accuracy, precision, recall, F1-score, and computation time were used to analyse the datasets. Based on the results achieved with the new dataset, the Random Forest, Support Vector Machine, Decision Tree, and K-NN achieved a 100% rate of 100% on the accuracy, precision, recall, and F1 score in a shorter computation time. With the open-source dataset, Random Forest, Decision Tree, and K-Nearest Neighbor achieved 100% accuracy.

Chapters

Downloads

Front Pages

Published

February 2, 2025

Categories

Copyright (c) 2025 Chisom Elizabeth Alozie

Details about the available publication format: E-Book

E-Book

ISBN-13 (15)

978-93-49307-78-0

Details about the available publication format: Book (Paperback)

Book (Paperback)

ISBN-13 (15)

978-93-49307-68-1

How to Cite

Alozie, C. E. . (2025). Analysing Cloud DDoS Attacks Using Supervised Machine Learning. Deep Science Publishing. https://doi.org/10.70593/978-93-49307-78-0
Download Citation