Intrusion Detection Systems: Enhancing Real-Time Network Threat Monitoring Using AI

Abstract

The failure of the traditional intrusion detection systems has prompted the search for better options. AI-based detection systems have proved superior to the traditional intrusion detection methods. Much research has been done on these aspects. This systematic review aimed to review the current trends of research on AI-IDS for network monitoring to enhance intrusion detection. Google Scholar was used for identifying the relevant papers with appropriate search terms. The identified papers were screened, and the most suitable 20 papers were selected using the PRISMA process flow diagram. The selected papers were described in the results section and thematically analysed and their quality rated in the discussion section. There is no doubt about the superiority of AI-based intrusion detection systems. Only the components of such systems differ depending on the target of the detection system, like cloud computing, IoT or internal structures and operating systems. The challenges to the implementation of AI-based intrusion detection systems in organisations have been identified. Solutions to these have been suggested. However, how many of these solutions have been implemented successfully by any organisation is unknown. Some case studies on big and small organisations can enlighten us on this aspect. Some limitations of this review and the scope for future research have been presented.