Implementation of Parallel BACON-MVV method based on Data Decomposition in Intrusion Detection System

In Computer Security area, Intrusion Detection System (IDS) plays important role in detecting any kinds of network attacks. Denial of Service (DoS) and Probing attacks are common detectable intrusions that are frightened by most network users since the final result of these attacks is collapsing the network. Our previous research has proposed a robust statistical method, the BACON-MVV method, that provides 100% accuracy in detecting patterns of DoS and Probing attacks, inspite of the training sets used contains suspicious packet traffic called outliers. One problem not yet being addressed by previous research was the processing time taken as the packet traffics to be analysed for detecting any intrusion grows bigger. In this paper, we propose a Parallel BACON-MVV method based on Data Decomposition to be implemented in IDS. Experiment using our own generated simulation datasets shows that this proposed method runs significantly faster than its serial version.