Emergency Vehicle Identification for Internet of Vehicles Based on Federated Learning and Homomorphic Encryption
Siyuan Zeng, Bo Mi, Darong Huang
Abstract
With the development of the Internet of Vehicles (IoV), its application has attracted wide attention. Emergency vehicles often have trouble moving in traffic. Therefore, the classification of vehicles into emergency and non-emergency categories is conducive to the development of IoV applications such as emergency rescue services, intelligent traffic management and autonomous driving systems. At the same time, its data is very sensitive in terms of data privacy and security issues. Federated learning, as a framework of machine learning, can be used to improve the privacy and security of data. The trained data is distributed on multiple machines to cooperate with each other for learning. In the process of federated learning, the model needs to be uploaded and downloaded. In order to ensure that the information of the model is not leaked, homomorphic encryption is used to encrypt the model to protect the information of the model. This paper presents a federated learning algorithm for IoV data privacy protection based on homomorphic encryption.