Intelligent Detection of Distributed Denial of Service Attacks: A Supervised Machine Learning and Ensemble Approach

Abstract

The rapid growth of IoT, smart devices, and 5G networks has increased the prevalence and complexity of Distributed Denial of Service (DDoS) attacks, posing significant challenges to cybersecurity. The objective of this research is to develop an effective method to detect and prevent DDoS attacks, thereby safeguarding communication systems from such threats. The proposed "Intelligent Distributed Denial of Service Attacks Detection (IDDOSAD) Approach" utilizes supervised machine learning algorithms, including Random Forests, Decision Trees, K-Nearest Neighbor, XGBoost, and Support Vector Machine, along with ensemble learning to enhance detection accuracy. the model development process consists of data collection, pre-processing, splitting into training and testing sets, selecting prediction models, and evaluating their performance. Evaluated on a dataset of 11,423 instances, the IDDOSAD approach demonstrated promising results, with accuracy ranging from 92% to 100% for the time series dataset. In conclusion, the IDDOSAD approach effectively detects and mitigates DDoS attacks, offering a reliable solution to protect communication systems against this growing cybersecurity threat.

Keywords

• Decision tree
• Random Forest
• DDoS attack
• Anomaly detection
• Ensemble Learning
• Machine learning