Scheduling Algorithms for the Aperiodic Tasks in Embedded Real-Time Computing Systems

The paper deals with the problem of task scheduling in embedded real-time computing systems, specially aperiodic tasks scheduling with soft deadlines. Algorithms for aperiodic tasks scheduling must satisfy hard deadlines of aperiodic tasks and anticipate good average response times of aperiodic tasks with soft deadlines. There are two basic approaches to servicing aperiodic requests with soft deadlines - background processing and polling tasks, the characteristics of which are improved by algorithms of a priority exchange server and a deferrable server. Better results in scheduling are obtained by using the sporadic server algorithm. Basic principles are given, applicabillity of procedures and improvements is evaluated through graphic presentation of scheduling.