Improvement of the Dynamic Priority Scheduling Algorithm Based on a Heapsort

The earliest deadline first (EDF) scheduling algorithm is a typical representative of the dynamic priority scheduling algorithm. However, once the system is overloaded, the deadline miss rate increases and the scheduling performance deteriorates sharply, which causes a reduction in system resource utilization. To overcome this problem, we proposed an improved dynamic priority scheduling algorithm based on heap sorting. The task deadline, task value, energy consumption, and other parameters were introduced. The fuzzy analytic hierarchy process (FAHP) and the value density method were then used to determine the comprehensive priority of tasks. A heapsort algorithm with a lower time complexity was used to sort the comprehensive priority index so as to reduce the sorting overhead of the system. The system sorting overhead was then introduced to improve the decision condition of the priority scheduling subset and expand the schedulable range of the algorithm. The experimental results showed that the improved method reduced the deadline miss rate by an average of 0.1789, which improved the scheduling performance of the algorithm. The optimized scheduling algorithm can be applied to industrial control to improve system efficiency, and reasonable resource scheduling can reduce data center costs.