基于组合优先级的自适应实时调度算法研究

在嵌入式实时操作系统中,由单个特征参数作为实时任务优先级的调度依据,并不能较好地描述系统中任务的关键性和紧迫性.提出一种基于组合优先级的自适应实时调度算法(SREDF),综合任务的截止期和CPU运行期设计任务的优先级,使截止期越早且CPU运行期越短的任务拥有最高优先级.处理器能有效地调度相同截止期的实时任务,并提前分析和预测任务能否完成.实验表明,该算法降低了任务调度的截止期错失率(MDP),并提高了任务调度的速度和效率.

Self-Adaptive Real-Time Schedule Algorithm Based on Combinational Priority Strategy

In the embedded real-time operating system,it is difficult to describe the importance and urgency of the task well by regarding single parameter as task priority of the schedule algorithm.A self-adaptive real-time schedule algorithm based on combinational priority strategy (SREDF) is proposed,which combines with the deadline and the runtime period,and sets the task which possesses the earliest deadline and the shortest of the CPU runtime the highest priority.The processor schedules those tasks with the same deadline,moreover,those tasks are analyzed and predicted in advance to judge whether they could be finished before the deadline.The experimental results show that the proposed algorithm decreases the missed deadline percentage(MDP) and improve the rate and efficiency of the task scheduling.