Effective Authentication Mechanism for Vehicular Fog Infrastructure

Abstract

Recently, a tangible vehicular fog computing (VFC) platform based on oneM2M, called oneVFC, has been proposed. This platform enabled federated AI learning, but did not include sufficient attention to security. Therefore, this paper proposes an effective authentication and key agreement protocol based on the Elliptic curve cryptography (ECC) to enhance privacy preservation in an authenticated way in order to ensure safer AI training. Experimental results show that the proposed authentication mechanisms make reliable transmissions in highly-mobile vehicular networks possible, resulting in a desirable quality of service (QoS). The performance and security strength of two authenticated encryption modes of AES, GCM and CCM, are also compared and evaluated.