A critical review on intrusion detection systems in IoT based on ML approach:A Survey

Abstract

Internet of Things (IoT) is extensively applied in a number of domains such as environmental monitoring, industrial processes, home automation, healthcare etc. due to numerous advantages associated with it. However, it is challenging to overcome the security issues and ensure privacy in IoT environment. Intrusion Detection Systems (IDSs) is an important phenomenon that secures the networks and information systems. But conventional IDS techniques cannot be applied in IoT because of limited resources and heterogeneity since IoT has specific features such as protocol stacks, standards and resource-constrained devices. A survey is conducted in this paper upon IDS-based studies focusing IoT. The objective is to find the trends, assess the issues and explore future research opportunities. Further, the study also discusses about different classification methods for IDSs, possibilities for every attribute and analyzes the IDS methods. A taxonomy was proposed to classify the papers based on validation strategy, security threat, IDS placement strategy and detection method.