AI-Powered Network Intrusion Detection via Packet Sniffing
No Thumbnail Available
Date
2025-09-04
Authors
Ahmad Sawafta
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
As networks become increasingly complex cyberattacks such as port scanning, distributed denial-of-service (DDoS) attacks, and injection-based vulnerabilities, persistently challenge the CIA of data. Traditional signature-based Intrusion Detection Systems (IDS) are often inadequate against novel attack vectors, generating high false positives and struggling with scalability. This paper presents the design, implementation, and comparative evaluation of a machine learning-based IDS that performs real-time traffic analysis through packet sniffing. The system captures network packets, preprocesses them to extract salient features, and employs a dynamic, multi-model analysis engine. We rigorously evaluated several machine learning algorithms, including Logistic Regression, Support Vector Machines (SVM), on a mixture of normal and malicious traffic datasets. Logistic Regression and SVM offered an optimal balance of high accuracy, low computational overhead, and minimal false positives, making them exceptionally suitable for real-time deployment. The proposed system demonstrates its scalability and adaptability for modern network security.
Keywords: Network Security, Intrusion Detection System (IDS), Machine Learning, Packet Sniffing, Real-time Analysis, Logistic Regression, SVM