Performance Evaluation of Dynamic Round Robin Algorithms for CPU Scheduling

Abstract

The performance of an operating system (OS) is affected by the algorithm policy that is used by a CPU to schedule the running processes. Thus, a better CPU scheduling algorithm results in faster OS performance using minimal resources over small amounts of time [11]. For that reason, many algorithms were proposed and implemented to enhance the performance of CPU scheduling. Round Robin is considered an efficient and fair algorithm because all processes are given the same amount of time quantum. However, its efficiency depends on the selected time quantum. In this paper, we present a comparative study of four different Round Robin algorithms namely: Adaptive Round Robin Algorithm, Best Time Quantum Round Robin CPU Scheduling, Optimal Round Robin Scheduling Using Manhattan Distance Algorithm, and Improved Round Robin Scheduling Algorithm. We compare these algorithms in terms of four performance factors including: Average Waiting Time (AWT), Average Turnaround Time (ATT), Average Response Time (ART) and Number of Contexts Switching (NCS). The simulation results show that both Adaptive Round Robin and Optimal Round Robin Scheduling Using Manhattan Distance algorithms are more efficient to be adopted as they recorded the minimum values of performance factors.