Search acceleration in preprocessing mechanism of network intrusion detection systems using graphics processors

Security is one of the crucial issues in today's world. There are several tools for providing security for computer networks. Intrusion Detection Systems (IDS) is one of the most widely used and effective tools in creating a safe environment for computer networks. In recent years, these systems have had significant progression while they still suffer from limitations against high traffic. Against high traffic, network intrusion detection system (NIDS) becomes a bottleneck and lowers the performance of computer network. In this paper, we employ the latest techniques and methods parallelism to achieve a higher performance of intrusion detection systems. In the current study, we implement search mechanism in the preprocessing of intrusion detection systems on graphics processors and achieved desirable performance. Several parallelism schema and examined for corresponding search algorithm and the final result shows more than 100x improvement in search algorithm speed.