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.

中文翻译：

---

物联网异常检测的最新进展：现状、挑战和前景


本白皮书对物联网 （IoT） 的异常检测进行了全面调查。异常检测在 IoT 中带来了许多挑战，应用范围广泛，包括入侵检测、欺诈监控、网络安全、工业自动化等。网络安全分析和研究人员已经受到了高度关注，尤其是网络中的异常检测，这在网络安全中至关重要。及时发现网络异常至关重要。由于各种问题和资源受限的特点，传统的异常检测策略无法在 IoT 中实现。因此，本文试图重点介绍各种最新的技术来检测 IoT 及其应用中的异常情况。我们还介绍了 IoT 架构的多个层的异常情况。此外，我们还讨论了多个计算平台，并强调了异常检测的各种挑战。最后，提出了这些方法的潜在未来方向，从而引发了各种开放的研究问题。通过这项调查，我们希望读者能够更好地了解异常检测以及该领域的研究趋势。