A three-phase task scheduling scheme in a hard real-time distributed environment

Abstract

The incentive of the local task scheduling scheme developed here is to guarantee a maximum number of unpredictable critical tasks locally. In order to guarantee the execution of a critical task within a requested time frame, a Safety Time Factor (STF) is defined based on the estimated worst-computation time of the task and added to its execution time. A three-phase local task scheduling is designed to take advantage of accumulative safety times and provide alternatives for scheduling the critical tasks locally. At each phase a different class of critical tasks is scheduled. The performance of the model has been ested using the most recent task scheduling scheme as a baseline model.