Energy-Efficient Partitioned-RM Scheduling for Shared Resources Imprecise Mixed-Criticality Tasks

Abstract

Shared resources and energy consumption are important factors to consider in the design of mixed-criticality systems. Existing works have studied these two factors separately. In this article, we simultaneously focus on shared resources and energy consumption on multiprocessor platforms. Firstly, we address the problem of energy-aware scheduling for the fixed-priority imprecise mixed-criticality tasks with shared resources and propose a schedulability test based on the Multiprocessor Priority Ceiling Protocol for a given task-to-processor mapping. Secondly, we calculate the energy-efficient speed of each processor based on the schedulability test and propose the corresponding task-to-processor mapping algorithm, called IMCPA. Finally, we conduct experiments on a real-world case and synthetic tasksets. The experimental results show that IMCPA can improve the schedulability ratio by about 13.76% and save energy consumption by about 34.89% compared to the existing algorithms.