Random Particle Swarm Optimization (RPSO) based Intrusion Detection System

Intrusion detection is a challenging area of research. As now there are several research work are already done and the result improvement is in progress. In this paper a hybrid combination of association rule mining and random particle swarm optimization (RPSO) has been applied. This approach is applied on NSL-KDD dataset. A limit set is provided by our framework which will be adapted as per the user choice to select the set of data for use. Our approach successfully differentiates the normal and attack node. Then we have applied a recheck frame for the normal node for finding the suspicious node. Then by the help of association rule associated values are passed for the next procedure. Then we apply RPSO to check the boundary value for the possible type of intrusion detection. If it is passed the velocity value then it will be listed in the attack type. Finally based on the attack category of Denial of Service (DoS),User to Root (U2R),Remote to User (R2L) and Probing (Probe) attacks are classified. The results of our method shows improvement in detection different type of attack in comparison to the previous method.