Recent advances in anomaly detection in Internet of Things: Status, challenges, and perspectives

Abstract

This paper provides a comprehensive survey of anomaly detection for the Internet of Things (IoT). Anomaly detection poses numerous challenges in IoT, with broad applications, including intrusion detection, fraud monitoring, cybersecurity, industrial automation, etc. Intensive attention has been received by network security analytics and researchers, particularly on anomaly detection in the network, deliberately crucial in network security. It is of critical importance to detect network anomalies timely. Due to various issues and resource-constrained features, conventional anomaly detection strategies cannot be implemented in the IoT. Hence, this paper attempts to highlight various recent techniques to detect anomalies in IoT and its applications. We also present anomalies at multiple layers of the IoT architecture. In addition, we discuss multiple computing platforms and highlight various challenges of anomaly detection. Finally, the potential future directions of the methods are suggested, leading to various open research issues to be analyzed afterward. With this survey, we hope that readers can get a better understanding of anomaly detection, as well as research trends in this domain.