Model of preemptive embedded systems for optimizing real-time performance

Abstract

By analyzing preemptions in embedded real-time systems adopting RM scheduling policy, created a preemption model of periodic real-time tasks. This model described the relationship between overheads incurred by preemptions and the properties of real-time tasks, such as priorities, periods, execution times, deadlines, release times etc. It also showed the relation between the schedulability of the task set and properties of its real-time tasks. Based on the model and taking the idea of parasitism in biology, this paper presented a method of optimizing real-time performance of the embedded systems under the RM scheduling mechanism. This method can reduce preemptions or change the relations of preemptions, decrease overheads to improve real-time performance by adjusting release times of tasks. And then the optimization method underlying the model is evaluated by experiments.

Keywords

• embedded systems
• release time
• real-time performance
• performance optimization
• preemption model
• evolutionary programming