A Semantic Review on Challenges, Trends towards Defensive IDS in Internet of Things

Abstract

Securing data over the network is one of the major issues among network users. Therefore, intrusion detection systems (IDS) were developed that become an important tool in security mechanism for Internet of Things (IoT). Over the years, a significant number of IDS have been developed; however, choosing the best can be tough sometimes. In this paper, different IDS developed over the last few years have been reviewed. The major focus of this review is to analyze the importance of IDS in computer networks and IoT, issues and possible solutions to overcome these challenges. In addition to this, role of machine learning (ML) and deep learning (DL) algorithms in detecting intrusions is also analyzed and discussed. Moreover, the key findings after reviewing the literature study are also mentioned in this work in which we find that majority of the researchers are moving towards the DL based algorithms for detecting intrusions because of their ability to handle large datasets. Therefore, by selecting a suitable method will enhance the detection rate.