Development of Intrusion Detection & Prevention System (IDPS) Using SVM Model with New Kernel Function for Security Solutions

Abstract

In digital era, every aspect has an online presence. Almost all the activities whether communication, financial transactions, shopping, social networking etc are carried out digitally. All these activities involve sharing of sensitive data and information, hence securing information has become inevitable for almost all agencies whether public or private. Confidentiality of information must be maintained for integrity of data, various studies have been proposed, to prevent the intrusion and cyber-attacks. Despite this, the hackers succeed in breaking the barriers and can access the data unauthentically. The proposed research study has produced a versatile Intrusion Detection and Prevention System (IDPS) to detect and prevent the intrusion of hackers in the web. To further enhance the performance of the proposed IDPS, SVM (Support Vector Machine) classification model, along with a linear kernel is used to classify the intruders, by using probability concepts in the SVM kernel function, which uses probability concepts to classify intruders and authenticated web users, suggested IDS is tuned for accuracy by using SVM classification model. An innovative SVM classification algorithm along with a new kernel function is used to develop and implement the new IDPS. This research work provides the overview on current active cyber-attacks (intrusions) as well as available intrusion detection and prevention methods.