AI-Driven Detection of Network Traffic Anomalies: A Case Study with OMNeT++
Year: 2025
Type: article
Abstract: This study explores machine learning-based anomaly detection for Distributed Denial-of-Service (DDoS) attacks using the OMNeT++ simulation platform. Network traffic features such as packet size, ports, and transmission frequency were collected to train Logistic Regression, Decision Trees, MLP, and Random Forest models.The Random Forest model achieved the highest accuracy (94%) and ROC-AUC score (0.927), demonstrating its effectiveness in intrusion detection. The study emphasizes data preprocessing, feature selection, and class balancing for improved model reliability.Future work will focus on real-time detection, adaptive learning, and expanded simulations to enhance network security
Source: Figshare
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6902414512
FWCI:
Cites:
Cited by:
Related to: