Detection Method for Software Defined Network Intrusion

Abstract

There is a great deal of interest in using Intrusion Detection Systems (IDSs) as a key component of system defense at the moment. To secure a network IDS analyzing network traffic from a location on the network or computer system. It is very difficult and time-consuming to distinguish between network traffic that is intrusive and normal. An analyst must examine that much and such a wide range of data. It is necessary to determine the sequence in which intrusions have occurred on the network connection. Current network traffic activity must be reflected by a way to detect network intrusions. Using genetic algorithm and machine learning approaches, a novel approach was developed to uncover intrusion characteristics for intrusion detection systems. The rules are generated using a genetic algorithm for decision trees. Rules can be applied to determine intrusion characteristics, and then implemented into a genetic algorithm for protection. In this way, in addition to identifying the presence of an intrusion, one may stop the incursion by rejecting it.