Network Intrusion Detection with Wiener Filter-based Agent
Mouhammd Sharari Alkasassbeh · 2011
Over the years, network attacks have evolved into more complex and sophisticated methods for network intrusion. As network domains become bigger in terms of size, the number of external users increases and thus the systems become more open and subject to security threats. Thus, the intrusion detection system was introduced to automatically detect network attacks. Commercial solutions however are generally centralized and suffer from scalability problems when applied in large networks. For this reason, in this paper, the distributed model was adopted to eliminate the scalability issues. Mobile agent combines with statistical methods based on the Wiener filter to collect and analyze the network data in order to detect anomalous behaviour in the network traffic. The algorithm was tested against four network attacks in both light and heavy traffic scenarios.