AI based Techniques for Network-based Intrusion Detection System: A Review

The internet has unlocked a whole new universe. It has no bounds and provides individuals with tremendous economic prospects all throughout the world. People can live better lives as a result of it. The internet has become one of the most important channels for communication. It has caused a massive range of information to be available online. This in turn has led to a lot of new threats coming into play, making it hard for network security to find breaches. An intrusion detection system (IDS) is a technology that scans network activity for unusual behavior and sends out alerts when it is found. It still has trouble detecting new intrusions, increasing the detection's precision and lowering false alert rates, despite the enormous efforts of the researchers. This research paper begins with a quick overview of IDS and its forms. We then go over several AI-based methods for Network-based IDS (NIDS), contrasting their advantages and disadvantages, while also determining evaluation metrics for each. Further discussing the various datasets used. We conclude our research by listing the research challenges along with the current and future trends.

I. Introduction

Cyberattacks are the most devastating and destructive form of modern warfare without weapons, resulting in the disclosure of sensitive personal and business data, disruption of critical operations, ongoing vulnerabilities, and unauthorized and illegal access to devices and software, all of which have a devastating impact on the economy of the nation [21]. These cyberattacks must be avoided. Intrusion-based cyberattacks are now commonplace, and early detection is critical. Currently, an IDS is still a must-have for protecting critical networks from outside intrusions[24].