A Hybrid Real-time Zero-day Attack Detection and Analysis System

摘要: A zero-day attack poses a serious threat to the Internet security as it exploits vulnerabilities in computer systems. Attackers take advantage of unknown nature and use them conjunction with highly sophisticated targeted attacks achieve stealthiness respect standard intrusion detection techniques. Thus, it's difficult defend against such attacks. Present research exhibits various issues is not able provide complete solution for analysis This paper presents novel hybrid system that integrates anomaly, behavior signature based techniques detecting analyzing real-time. It has layered modular design which helps high performance, flexibility scalability. The implemented evaluated metrics like True Positive Rate (TPR), False (FPR), F- Measure, Total Accuracy (ACC) Receiver Operating Characteristic (ROC) curve. result shows rate nearly zero false positives. Additionally, proposed compared Honeynet system.