Smart Vehicles Recommendation System for Artificial Intelligence-enabled Communication

The Internet of Things (IoT) and the Electric Vehicle (EV) industry are critical to the rapid growth of consumer-centric Internet of Vehicles (IoV) services to facilitate 6G-enabled vehicle communication, which offers numerous advantages. Among the applications of the Internet of Vehicles (IoV), a recommendation system is introduced to identify nearby charging sources while preserving user privacy, a crucial aspect of the IoV framework security. Determining which charging sources to suggest to an EV (as consumer electronics) is a challenging issue, given the numerous potential recommendations available. This paper introduces a secure recommendation system for EV consumer electronics, considering both fixed and mobile charging locations, focusing on optimizing the well-being of EV consumers as well as owners. Unlike traditional methods that involve sharing data directly among data holders during model training, our model employs a secure vertical Federated Learning (FL) approach, ensuring that data from EVs and charging sources remains within their respective platforms. To enhance model efficiency and address communication-related concerns, we employ fog-based data aggregators with 6G network communication, responsible for transmitting locally computed training parameters instead of conventional centralized architectures. Simulation results from our recommended system show a more optimal distribution of EVs within designated areas.