As automotive electronic systems become increasingly complex, the demand for
robust data security and privacy protection mechanisms has grown significantly.
The AUTOSAR (Automotive Open System Architecture) standard has emerged as a
widely adopted framework in the automotive industry due to its strong support
for interoperability, functional safety, and cybersecurity. Within the AUTOSAR
Classic Platform (CP), the Crypto Stack Service as a core component that enables
critical security functionalities such as encryption, decryption, digital
signature verification, and key management.
However, the deployment of the Crypto Stack across heterogeneous Electronic
Control Units (ECUs) introduces a series of technical challenges. These
challenges stem primarily from variations in hardware resources, differences in
operating system implementations, and inconsistencies in software execution
environments. As a result, issues such as architectural compatibility, task
scheduling efficiency, and secure communication between modules must be
addressed for successful integration.
This paper presents a systematic adaptation framework for the AUTOSAR Crypto
Stack, focusing on three key layers of the software architecture: the Operating
System (OS), the Runtime Environment (RTE), and the crypto driver abstraction.
The proposed solution includes optimized task scheduling strategies,
standardized RTE service encapsulation, and a dynamic dispatch mechanism for
coordinating software- and hardware-based crypto processing.
To validate the proposed adaptation strategy, a real-world prototype was
developed using the NXP S32K148 platform. The system was tested through the
generation and verification of MAC, simulating realistic automotive use cases.
Experimental results demonstrate that the solution meets the stringent real-time
and security requirements of automotive systems, providing valuable insights for
the secure deployment of Crypto Stack in modern vehicles.