Intelligent Internet of Things Attack Detection: Novel Approaches and Technologies

The Internet of Things (IoT) influences numerous facets of our daily lives, and artificial intelligence (AI)-based programs and applications are rapidly gaining popularity. Due to the vast adaptability of the IoT and the escalating data privacy concerns, classical AI algorithms may not be applicable in certain practical application domains. Federated learning (FL) has emerged as a viable technique for training distributed machine learning (ML) models while maintaining data privacy and security. In recent years, it has gained attention in the context of IoT security, particularly for detecting attacks on IoT devices and networks. In this article, we examine ML and DL in the context of the IoT and IoT attack detection. Then, we evaluate the most recent FL applications in IoT networks and IoT attack detection, investigating their potential to deliver a variety of IoT services, including security and privacy. Our work on detecting FL-IoT attacks has taught us a lot, and at the end, we suggest some directions for future research in this quickly growing field.