Analytical Evaluation ofPriority-Based Real-TimeSchedulingAlgorithms forIndependentPeriodicTasks

Abstract

Scheduling is an inevitable part of any multitasking Real-Time Operating System. Thereal-time scheduling schemes must guarantee the execution of the tasks to complete inreal time, i.e. a task may have some completion deadline. Real-time schedulingalgorithms are different fromgeneral scheduling algorithms as they deal with the timeconstraints. Consequently, these real-time scheduling techniques need special attention togive.This dissertation focuses on the study of the two priority-based real-time schedulingalgorithms; Rate Monotonic Scheduling (RMS) and Earliest Deadline First (EDF). Thesealgorithms are analyzed in terms of processor utilization, average turnaround time,context switches, average waiting time, of the tasks for particulartask set.Moreover, the general scheduling algorithms like First Come First Serve (FCFS),Shortest Remaining Time Next (SRT), Shortest Task Next (STN) and Round Robin (RR)are explored for evaluating the real-time scheduling algorithms. Though, they lack mostimportant aspect of the real-time scheduling i.e. time constraints (e.g. deadlines), otherstatistics are analyzed to compensate with the real-time scheduling algorithms.The results of this dissertation provide a foundation for inferring optimality andfeasibility of priority-based real-time scheduling theory. It is expected that these resultswill aid the system designers in the development of predictable real-time systems.