Enhancing response times of end-to-end tasks using slack of local tasks

Abstract

This paper aims to enhance the response times of soft real-time end-to-end tasks while guaranteeing deadlines of hard real-time local tasks. Since the end-to-end task is comprised of multiple subtasks, we derive the intermediate timing attributes of subtasks taking into account precedence relations, and find the highest possible priorities of subtasks using the slack of hard real-time local tasks. The final result is a set of periodic tasks, which can be scheduled according to a fixed-priority scheme.