Research On The Scheduling Policy Of Embedded Operating System

The scheduling policy is a key part of the embedded operating system. It is an important research field in the embedded operating system that how to schedule tasks and make every task to be finished in time. The simplification and high efficiency of the scheduling is significant for improving the whole capability of the embedded system.In the dissertation, the schedule algorithm for the embedded operating system was researched. And the scheduling policy of the RTOS uClinux was studied. The task-resource mold based on the periodical task was then developed. Many time parameters were imported .So the real-time system so that is easy to make model. The model had been ready for the schedule algorithm in order to analyze and experiment.The Priority-driven Scheduling Algorithm was based on the task-resource model. In the algorithm, RTOS assigned weight to every resource and built evaluation function. The algorithm gave a better result for the problem about reversal and improved stabilization and real-time performance of Embedded RTOS effectively.And then, we put forward the scheduling frame for the embedded operating. The frame was scalable, mixed, stability and reliability. The scalable frame sustained many kinds of scheduling arithmetic and service. The frame is applicable to the embedded system especially.We implemented the Priority-driven Scheduling Algorithm Based on Weight in the MicroC/OS-II2.52. MicroC/OS-II is free for teaching and research. So, this work will be representative.At last, A Priority-driven Scheduling Algorithm Based on Weight was embedded in the operating system. I pricked off the up to the operating system MicroC/OS-II2.52 to SamArmDvk4.0. So the algorithm was come true and validated.The task-resource mold was contained the attribute of time and resource. So it was propitious to analyze the embedded system. A Priority-driven Scheduling Algorithm Based on Weight is excellent capable of system resources and has stability for the embedded system. The scheduling frame has very good performance for the complex and real-time embed system.