With the rapid development of the Internet of Things (IoT), how to protect users’ privacy security in the IoT environment has received extensive attention. Federated learning allows IoT devices to perform local training and only upload model parameters, avoiding the direct transmission of raw data, which provides a new solution for enhancing IoT privacy security. Many studies have tried to integrate federated learning with other technologies to further improve privacy protection. This paper summarizes various technologies used in recent studies to protect privacy security in different IoT scenarios with federated learning, including encryption techniques, secure model aggregation, and the integration of distributed trust mechanisms. In addition, this paper also introduces the applications of federated learning in various IoT scenarios, including industrial IoT, healthcare, and energy management fields. The paper also provides future prospects for research on using federated learning to protect privacy security in the IoT. Furthermore, it explores potential advancements in combining emerging technologies such as blockchain and differential privacy to achieve more efficient and secure privacy protection mechanisms in IoT environments.
