Improved Round Robin CPU Scheduling Algorithm with Different Arrival Times Based on Dynamic Quantum

Abstract

Modern operating systems are based on the principle of time-sharing in executing simultaneous operations. Determining the length of the time slice, and the times when processes arrive at the ready queue are problems that affect metrics as the average waiting time (AWT), average turnaround time (ATAT), response time (RT) and the number of context switches (NCS) of the time-sharing round robin RR algorithms. The research aims to propose an algorithm that achieves a short waiting time while maintaining a reasonable response time, which is the most important characteristic of time-sharing algorithms. The Different Arrival-Dynamic Quantum Round Robin (DADQRR) algorithm bases its work on different parameters to adjust the time slice value dynamically. The algorithm has been compared to three other algorithms that are similar in terms of dealing with different arrival times, namely AN, MARR, RR. The algorithm outperformed the three algorithms at range from 6.155% to 31.409% in term of AWT. It achieved an outperformance of 5.924% to 30.850%, considering the TAT. The ranges of outperformance values resulted from the difference in the ranges of arrival times, as well as in the ranges of burst times.