An Integrated Design Method of Task Priority

A scheduling algorithm based on priority table design is presented in this paper. Any two characteristic parameters (e.g., relative deadline and slack) of a task are combined to design its priority table so that the deadline is nearer or the slack is shorter, the priority is higher. The priority of a task is uniquely determined by its relative deadline and slack. For any task, its unique priority can be obtained by using Lagrange interposing algorithm on the designed priority table. Compared with the classical EDF and LSF policies, simulated results show that the proposed algorithm improves the efficiency of task scheduling, i.e., to designate the priority of a task, increases the succeed ratio of task scheduling, and decreases the missed deadline percentage. The proposed algorithm can be applied to dynamically schedule real-time tasks in real-time systems.