A distributed network intrusion detection system architecture based on computer stations using GPGPU

Abstract

This article deals with the proposal of distributed parallel network system architecture for intrusion detection. The aim is to design model of architecture that will be used for searching, detecting and analyzing intrusions. This proposal of detection system in network will use enormous performance of the GPGPU high-parallel data computing technology which can substitute the most complicated part of the intrusion detection process in NIDS. The model is highly scalable and allows the usage of several most recent trends in the IDS field with GPGPU support. Due to the fast attacks reaction system analyzes all the data in different formats in whole set of attacks and system alerts.