An Effective Anomaly Intrusion Detection Using Statistical Change Point Detection

Abstract

Understanding the nature of intrusion attacks is critically important to the development of effective counter measures to anomaly traffic detection problem. Anomaly intrusion traffic attacks combined with traditional network intruders became most serious threats to network security. The existing work monitors available traffic attacks and take appropriate action to mitigate them, before they have had much time to propagate across the network. The proposed working model of statistical traffic anomaly detection method is carried out on the principle traces of non intrusive packet header data with quick detection rate. Traffic is monitored at regular intervals to obtain a signal that can be analyzed through statistical techniques and compared to historical norms to detect anomalies (change detection). The proposed methodology of anomaly intrusion traffic detection envisions statistical change detection theory for real-time data source extracted from Net Con server (Internet Service Provider popularly running at Erode Region). The experimental results suggest little use of address spoofing by attackers, which imply that such attacks will be invisible to indirect backscatter measurement techniques. The proposed traffic anomaly intrusion detection provides an improvement of 12% average through put compared to the existing ones. The propagation delay metric shows a reduction of nearly 9% with other methods of anomaly intrusion detection.