Latency-Aware Round Robin Variants for Heterogeneous Workload Allocation

Abstract

Operating Systems (OS) performs a variety of tasks, with scheduling being a fundamental function. Scheduling involves making decisions on how to allocate resources among different processes in tasks in order to maximize performance metrics. Some of the popular scheduling algorithms are First Come First Serve (FCFS), Shortest Job First (SJF), Shortest Remaining Time (SRT), Priority Scheduling, and Round Robin (RR), which exhibit differences in performance depending on the circumstances. The primary goal of a scheduling algorithm is to increase the effectiveness, speed, and fairness of the system. Considering these parameters, to improve and be suitable scheduling algorithm, studies have proposed hybrid algorithms. To understand the performance trade-offs between these hybrids, this paper compares different RR-based hybrid algorithms namely, Priority Based Round Robin (PBRR), Round Robin and Short Job First (RRSJF), Efficient Shortest Remaining Time Round Robin (ESRR). Results showed that RRSJF and ESRR performs better than the remaining hybrid algorithm in terms of Average Waiting Time (AWT) and Average Turnaround Time (ATT) with 17.2ms and 13ms AWT 13 and 26.0ms and 21.8ms ATT, respectively.