Efficient Multi-Resource Scheduling and Allocation in Multi-Core Automotive Control Systems

Develop a multi-resource collaborative scheduling and allocation framework for multi-core automotive control systems that efficiently schedules tasks and properly allocates multiple shared resource types, including memory, bus, and I/O devices, in order to address the increased coordination and management complexity across these resources.

Background

The paper proposes BR-WFD, a task allocation algorithm that integrates blocking time estimation and resource similarity analysis to improve schedulability under shared resource contention in multi-core real-time vehicle control systems. The approach targets synchronization blocking effects under MPCP, focusing on minimizing global blocking by co-locating tasks with shared resources and prioritizing tasks with high synchronization sensitivity.

In the conclusion, the authors note that future multi-core automotive control systems will involve a broader set of shared resources (e.g., memory, bus, I/O devices), whose coordination and management will be more complex. They explicitly state that achieving efficient scheduling and proper allocation across multiple resource types remains unresolved, motivating the need for a multi-resource collaborative scheduling framework.