PBatch: Pseudonym Certificate Batch Authentication With Generative AI-Based Cache for Cooperative Intelligent Transportation Systems

Abstract

Authentication and revocation are the key mechanisms to ensure the security of the Cooperative Intelligent Transportation System (C-ITS). C-ITS relies on the Vehicular Public Key Infrastructure (VPKI) for anonymous authentication and device revocation. Several works complemented the VPKI-based authentication and revocation process. However, several security and performance issues exist in both mechanisms. This article presents PBatch: Pseudonym Certificate Batch Authentication based on Distributed Ledger Technology. PBatch addresses challenges specific to the authentication and revocation process to achieve 1000 authentications per second. PBatch relies on the concept of batching pseudonym certificates by offloading heavy validation operations such as certificate chain and revocation status validation to local edge servers. This enables vehicles to validate a batch of pseudonym certificates with a fixed number of verification operations, thus simplifying the authentication of the pseudonym certificate at the end devices. Furthermore, a caching-based message authentication mechanism is introduced to validate a relatively larger number of safety messages. We also introduced a Generative Artificial Intelligence (GAI) based cache management mechanism for safety messages caching and fetching. Finally, experiments and security analysis are conducted to investigate PBatch performance and security. The results show that PBatch is more secure, feasible, and scalable than the leading VPKI-based authentication proposals.