Chi-Square Statistical based Technique for Intrusion Detection

Tools required for the security purposes are firewall, passwords, IDS, IPS for the detection of anomaly and prevent it from sending out the harmful traffic to the network. So, it is very necessary to examine the behavior of traffic that coming to the network and recognize the anomalous behavior. In this paper, statistical based chi-square method is used to detect the anomalous behavior and predict the intrusions by calculating the observed and expected frequencies. Setting of interval is difficult for the detection of anomaly but in our case we set interval according to the less variation in traffic. Traffic contains from the backscatter dataset. Chi-square method is good to detect the anomalous behaviors because it gives the Poisson’s distribution for the whole traffic on network. Large difference shows anomaly occurs.