Flexible and Dynamic Scheduling of Mixed-Criticality Systems.

Area of Science:

• Computer Science
• Embedded Systems Engineering
• Real-Time Systems

Background:

• Mixed-criticality systems integrate tasks of varying importance on shared platforms.
• Ensuring high-criticality task execution while maximizing low-criticality task service is a key challenge.
• Existing pessimistic scheduling strategies can lead to resource inefficiency and reduced performance for low-criticality tasks.

Purpose of the Study:

• To address resource waste and performance degradation in flexible mixed-criticality systems.
• To develop a dynamic scheduling scheme that improves resource utilization and task performance.
• To reduce system switching costs and unnecessary redundant resource expenditures.

Main Methods:

• Proposed an adaptive-service-level adjustment strategy for low-criticality tasks.
• Developed a flexible and dynamic mixed-criticality system scheduling scheme.
• Designed a system execution framework, scheduling algorithm, and dynamic maximum execution budget allocation strategy.

Main Results:

• The proposed adaptive strategy effectively solves resource waste issues from invalid compensation.
• The dynamic scheduling scheme significantly reduces redundant resource expenditures and system switching costs.
• Experimental results demonstrate improved performance for low-criticality tasks compared to existing methods.

Conclusions:

• The novel adaptive and dynamic scheduling strategies enhance the efficiency and performance of mixed-criticality systems.
• The proposed methods offer a more effective approach to managing tasks with different criticality levels.
• This research contributes to optimizing resource allocation and task scheduling in complex embedded systems.