Related Experiment Video
Updated: Jul 1, 2025
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
531
Novel Machine Learning Approach for DDoS Cloud Detection: Bayesian-Based CNN and Data Fusion Enhancements.
Ibtihal AlSaleh1, Aida Al-Samawi1, Liyth Nissirat1
1College of Computer Sciences and Information Technology, Department of Computer Networks, King Faisal University, Al-Ahsa 31982, Saudi Arabia.
Sensors (Basel, Switzerland)
|March 13, 2024
Summary
This study introduces a machine learning model for detecting Distributed Denial of Service (DDoS) attacks in cloud environments, achieving high accuracy. The advanced BaysFusCNN approach further enhances detection reliability for cloud security.
Area of Science:
- Cloud Computing Security
- Machine Learning Applications
- Cybersecurity Threat Detection
Background:
- Cloud adoption is widespread, with 95% of enterprises using cloud technology and 79% migrating workloads.
- Cloud environments face significant cybersecurity risks, including network vulnerabilities and Distributed Denial of Service (DDoS) attacks.
- Traditional Intrusion Detection Systems (IDS) have limitations in effectively detecting sophisticated DDoS attacks in the cloud.
Purpose of the Study:
- To develop an innovative machine learning approach for enhanced DDoS attack detection in cloud computing.
- To improve the accuracy and reliability of Intrusion Detection Systems (IDS) within cloud infrastructures.
- To address the limitations of existing methods for identifying and mitigating DDoS threats in cloud environments.
Main Methods:
- Proposed a Bayesian-based Convolutional Neural Network (BaysCNN) model for DDoS cloud detection.
- Utilized the CICDDoS2019 dataset with 88 features and applied Principal Component Analysis (PCA) for dimensionality reduction.
- Developed an enhanced Data Fusion BaysFusCNN approach incorporating Bayesian methods for uncertainty estimation and multi-source feature integration.
Main Results:
- The BaysCNN model achieved an average accuracy of 99.66% in detecting 13 multi-class DDoS attacks.
- The Data Fusion BaysFusCNN approach further improved accuracy to 99.79% across the same 13 multi-class attacks.
- The proposed models demonstrated significant enhancements in DDoS cloud detection accuracy and reliability.
Conclusions:
- The developed machine learning models offer a robust solution for detecting DDoS attacks in cloud environments.
- The findings provide valuable insights for creating advanced, reliable, and scalable machine learning-based IDS.
- The methodology empowers organizations to proactively mitigate cloud security risks and strengthen defenses against cyber-attacks.
