Indian Journal of Science and Technology

Abstract

Background/Objectives: In current days, administrating security effectively for computer resources has become a difficult task for the administrator. One of the security problems is Denial of Service (DoS), is a type of attack that tries to prevent legitimate users from accessing either the services or resources, by generating large number of artificial packets send towards the victim resource. In turn, the victim resource is unable to extend the service to legitimate user. To meet this type of attacks, numerous detecting and preventing systems have been proposed, but they were suffering from low detection accuracy and high false alarms. Methods/Statistical Analysis: A new computational technique was proposed to perform the classification task andextracting features from KDDCUP 99 datasets using genetic algorithm and Particle Swarm Optimization. This research work focuses on the identification of DoS attacks with high detection accuracy and less false alarms. Findings/Conclusion: In the proposed Intrusion Detection System model, attacks are identified by training the Particle Swarm Optimization classifiers with Genetic-Particle Swarm Optimization based on wrapper feature selection which is superior to those classical intrusion feature selection. The proposed work was implemented in MATLAB. The result shows high detection accuracy and fewer false alarms compared to the existing models.
Keywords: Denial of Service, Genetic Algorithm, Intrusion Detection System, Particle Swarm Optimization