📄 Abstract
With the rapid growth of cyber-attacks, traditional firewall systems that depend on fixed and manually created rules are no longer effective. This paper proposes an intelligent system for firewall rule optimization and anomaly detection using a hybrid approach that combines One-Class SVM and K-Means clustering. The system continuously monitors network traffic and learns normal behavior patterns. The One-Class SVM model is used to detect unusual or suspicious activities (anomalies), including unknown or zero-day attacks. At the same time, K-Means clustering groups similar firewall rules and network patterns to identify redundant, conflicting, or unnecessary rules. Based on this analysis, the system can automatically suggest better firewall rules, remove duplicate rules, and improve overall network security. This reduces manual work for administrators and helps the firewall respond faster to new threats. The proposed method improves anomaly detection accuracy, reduces false alarms, and enhances firewall performance. Overall, it provides a smart, automated, and efficient solution for modern network security systems.
🏷️ Keywords
📚 How to Cite:
Ismail Khan Mohammed, Srinubabu Sanagala , Somisetty Pavan Teja, Ikkurthi Bala Venkata Krishna, Yemineni Venkata Sai Rama Koteswara Rao , AN INTELLIGENT FIREWALL RULE OPTIMIZATION AND ANOMALY DETECTION FRAMEWORK USING HYBRID ONE-CLASS SVM AND K-MEANS CLUSTERING , Volume 11 , Issue 10, October 2026, EPRA International Journal of Research & Development (IJRD) , Pages: 208 - 213 ,