Performance evaluation of hybrid round robin algorithm and modified round robin algorithm in cloud computing

Abstract

The scheduling Round Robin (RR) is an impartial algorithm that schedules cloud resources by giving static time quantum to all processes. Time quantum selection is very crucial as it determines performance of algorithms. This research paper suggests an approach to improve RR scheduling algorithm in cloud computing by considering the quantum to be equal to burst time of start request, which dynamically vary after each execution of a request. And also, if the remaining burst time of CPU for currently executing process is lesser than time quantum, then the CPU will be allocated again to the executing process for rest of CPU burst time. MatLAb was used to implement the planned algorithm and benchmarked against MRRA available in literature. In comparison with the planned algorithm, Average Turnaround Time (ATAT) and minimal Average Waiting Time (AWT) was recorded. Based on the obtained simulated outcome, the planned algorithm should be preferred over modified round robin algorithm as it significantly improves the system efficiency.