The increasing number of vehicles being connected to the IoT, the integration between the IoT and vehicles has led to the emergence of the Internet of Vehicles (IoV) from the traditional Vehicular Ad-Hoc Networks. Internet of Vehicles (IoVs) is an open and integrated network characterized by collaborative environment data sensing, computing, and processing.
Artificial Intelligence (AI) based machine learning technologies have been widely utilized on the Internet of Vehicles (IoVs) and have shown significant advantages and efficiency for sharing and transmitting data among intelligent vehicles. Due to the enormous amounts of dynamic data generated exponentially in the IoV network, traditional machine learning models faces sufficient storage, computing resources, and data silos issue to store and process the massive amount of data by training the learning model at a central data server.
To combat this issue, Federated learning (FL) is a promising approach that enables privacy-preserving machine learning that can be implemented in the IoV to efficiently train the learning models by preserving the privacy of raw data and reducing the transmission overhead in wireless communications. FL empowers IoV to assist the vehicular service provider (VSP) and improves the global model accuracy for road safety and better profits for both VSP and participating smart vehicles (SVs).
In recent years, the development of IoV and service capability has spawned many applications involving driving safety applications, traffic efficiency applications, and entertainment applications. Nevertheless, Federated Learning on the Internet of Vehicles poses some major challenges, such as dynamic activities and diverse quality-of-information (QoI) from many SVs, VSP-s limited payment budget, and profit competition among SVs.