Privacy-Preserving Trust Management For Vehicular Communications and Federated Learning

Abstract

Cellular networking is evolving as a wireless technology to support a wide range of applications in vehicular communication. Cellular technologies enable vehicles to communicate with many applications to improve driving experience. All of these applications are not under the administration of the same authority. Secure Credential Management System (SCMS) is a Public Key Infrastructure (PKI) which provides certificates to vehicles to preserve vehicular privacy and supports many vehicular applications such as Basic Safety Messages (BSMs), misbehavior reporting, etc. On the other hand, privacy-preserving applications like federated learning (used to enable data-driven machine learning for better self-driving experience while protecting the vehicular data privacy) are outside of this SCMS-managed vehicular network and have separate PKI structure managed by a different authority. Since the trust management and PKI for vehicular vs. federated learning have been studied and developed separately, there is a need to establish trust between these two PKIs for secure communication between vehicles and Federated Learning Servers (FLSs). In this work, we demonstrate and analyze how these two orthogonal PKIs can establish trust with each other and thus, the end entities (vehicles and FLSs) can securely communicate with each other in an efficient manner to maintain vehicular privacy utilizing SCMS.