The detection method of identification terms for intrusion detection system

Abstract

Computer systems are increasingly exposed to threats, new types of which generate new types of cyberattacks on their resources. To increase the security level there were developed special systems focus on detecting abnormal condition in computer networks and formation of fuzzy measurement standards of the network parameters and the formation of heuristic rules for network activity assessment. The basis of these systems is the method of anomalies detection caused by cyber attacks. In this method, the detection process of terms identification is not accompanied by sufficient level of formalisation for its effective use. In order to eliminate this shortcoming, the detection method of terms identification, which is based on mathematical models and fuzzy logic methods is developed. The method is implemented through three basic stages: formation of multitude features; raiting the subsets of features; defining the number of term identification. It makes possible to search in a given linguistic variable the reference term according to which, with a help of heuristic rules, it can be possible to define the level of abnormal conditions specific to a particular type of cyber-attacks. This will increase the efficiency of the intrusion detection systems.