FPGA Acceleration and ARM Collaborative Optimization Testing of Multi Task Scheduling Algorithm Based on Zynq SoC
Lanlai Wang, Li Xu, Wei Liu, Xiaohai Hu, Yongchang Zhao · 2025
Existing multi-task scheduling algorithms face challenges in resource utilization and real-time performance. How to efficiently implement task scheduling to optimize system performance has become an urgent problem to be solved. First, a multi-task scheduling framework is designed to divide tasks into different priorities. Then, hardware acceleration of the scheduling algorithm is implemented on the FPGA, and pipeline processing and parallel computing are used to improve throughput. Next, the ARM processor is used for task management and scheduling control, a time-slice-based scheduling strategy is adopted, and dynamic priority adjustment is implemented. Finally, real-time communication between FPGA and ARM is established through an efficient data transfer mechanism (DMA). The experimental result in the paper shows a resource utilization of 74% achieved by the Method in 20 tasks, which far exceeded the 57% attained by the traditional methods. The multi-task scheduling algorithm based on Zynq SoC can boost scheduling efficiency dramatically, and in this way, provide individualized solutions for future embedded system designs.