Confidentiality, integrity, and non-repudiation of automobile event data: a TEE-based secure transmission scheme between automobile and cloud

Confidentiality, integrity, and non-repudiation of automobile event data: a TEE-based secure transmission scheme between automobile and cloud

연구 분야: Cryptography

학회: Cluster Computing

Modern automobiles are evolving into intricate cyber-physical systems, equipped not only with numerous internal sensors collecting a myriad of data but also interconnected to the Internet. This electronic data in automobiles plays a crucial role in automotive digital forensics. The Event Data Recorder (EDR), designed specifically to record accident-related data for motor vehicles, theoretically provides evidence admissible in court. However, this has heightened concerns about the confidentiality, integrity, and non-repudiation of such information. This paper introduces an accident-driven automotive EDR data secure uplink scheme to address these needs. We first analyze the EDR data uplink framework within the automotive Electrical/Electronic (E/E) architecture based on domain controllers. Subsequently, the EDR data is segmented into three parts, each undergoing sequential signing and encryption. Concurrently, digital envelope technology is employed to encrypt the session key for transmission to the cloud. Before submitting EDR data as evidence, we entrust a trusted third party to decrypt and verify the signatures; to ensure that the final evidence has not been tampered with. Simultaneously, all security-related operations operate within a Trusted Execution Environment(TEE), utilizing hardware isolation techniques to thwart malicious access controls. Security analysis and test results indicate that the proposed scheme can safely and efficiently transmits EDR data.