Literature Review of Machine Learning Models on Intrusion Detection for Internet of Things Attacks
Aminanto Achmad Eriza, M T Survadi · 2021 International Conference on Electrical, Computer and Energy Technologies (ICECET) · 2021
The Internet of things provides many useful applications for ease of living. Unfortunately, the IoT has resource-constrained and open vulnerabilities which are exploited by attackers to become their cyber army to do malicious activities. IoT devices are often exploited to launch denial of services attacks. One common countermeasure of the problem is an anomaly-based intrusion detection system. This kind of IDS typically leverages machine learning models to classify attacks. With the advance of underlying feature representation learning of deep learning models, IDS can distinguish between normal and malicious activities. In order to prevent disorientation for future researchers in this topic, we explore several state-of-the-art papers which leverage deep learning models for IDS in IoT network attacks. This paper focuses on one recent IoT attack dataset that was published in 2019. We listed researches that are leveraging the dataset. By discussing all related work using the dataset, we found the research direction in IDS on IoT attack research.