Modern vehicles increasingly rely on software-intensive functions, which continue to drive the growth and complexity of Electronic Control Units (ECUs). Conventional architectures based on single core or statically partitioned multicore microcontrollers are becoming inadequate for meeting emerging requirements such as deterministic real time performance, high communication throughput, synchronized time bases, and secure message handling. This paper introduces a software driven multicore consolidation approach that enables multiple ECU functions to be integrated on a single high-performance microcontroller while maintaining essential attributes of safety, determinism, and responsiveness.
The proposed architecture emphasizes the co-location of dependent software modules to reduce cross core communication overhead and enhance runtime efficiency. By leveraging AUTOSAR compliant multicore Basic Software (BSW) stacks, including Communication, NVM, STBM, Diagnostics, and OTA, the system supports improvements in timing accuracy, data throughput, and task synchronization. In addition, core-specific communication channel assignment, interrupt-based routing, HSM supported secure frame handling, and multicore enabled NVM with a master–satellite design collectively contribute to more efficient CPU utilization and reduced latency under high load communication scenarios.
The approach supports key industry relevant capabilities (e.g., runtime remote software updates, sensor fusion) and enables efficient handling of high-volume communication activity, making it adaptable for emerging multicore ECU architectures in next generation vehicles. The theoretical analysis presented in this paper describes how the proposed software driven consolidation can reduce system overheads, improve the organization of multicore workloads, and support more structured parallel execution.
The study further highlights that a well-designed multicore software strategy can serve as a practical and scalable direction for future automotive electronics, supporting higher levels of feature integration and communication flexibility, along with lower CPU load compared to traditional multicore ECU implementations. Considerations for future enhancements in software component grouping and communication stack adaptability are also discussed.