A system based on Naive Bayesian for Denial-Of-Service Attack detection

Abstract

Denial-of-service (DoS) attacks cause serious effect on systems. For most correct network traffic characterization, attack detection system uses multivariate correlation analysis (MCA). It Extract the geometrical correlations in between network traffic features. MCA based system enlightens the principle of anomaly based detection while attack recognition. MCA makes the situation easy for detecting known and unknown types of DoS attacks by simply observing the legitimate network traffic patterns. MCA uses Triangle Area Map (TAM) technique to speed up the Multivariate Correlation Analysis process. Proposed system can be evaluated by using KDD cup99 dataset. Naive Bayes (NBS) classifier is used as for attack detection. This algorithm addresses the problem of classifying the large intrusion detection dataset, which improves the detection rates and reduces the false positives at acceptable level in intrusion detection.It is probabilistic classifier which based on applying Bayes theorem.The proposed DoS attack detection system achieved highest accuracy as comparing to RBFN and IBK.99.96% accuracy is achieved by intrusion detection system.The Proposed detection system gives very low false positive Rate as about 0.002% which helps to increase the performance of detection System. As compare to RBFN and IBK, Naïve bayes classifier gives very low false positive rate, which helps to increase the performance of detection System. As compare to RBFN and IBK, Naïve bayes classifier gives very low false positive rate.