A new Improved Round Robin-Based Scheduling Algorithm-A comparative Analysis

Abstract

CPU or process Scheduling, which is an important part in any operating systems, allocates processes to the CPU in specific order to optimize some objective functions. The efficiency of any operating system relies strongly on the scheduling algorithms used. A number of scheduling algorithms exists. Among them, Round Robin (RR) is the most widely utilized algorithm. RR has proved to be effective in several types of operating systems, such as time sharing systems. This is due to the reasonable response time it gives. However, it suffers from some shortcomings such as high average turnaround time, high average waiting time as well as many context switches. Recently, several algorithms have been proposed to improve its efficiency, however, few studies were conducted to compare their efficiency when applied to dataset with different characteristics. The aim of this paper is three-fold: (1) presenting a survey of various RR based scheduling algorithms proposed and found in literature, (2) proposing a new RR based approach named, the Eighty-Five Percentile Round-Robin algorithm (EFPRR), to overcome the aforementioned issues, and (3) Conducting comparisons between eight RR based algorithms and the proposed approach using datasets with different characteristics. Extensive experiments have been done to test the proposed approach. Experimentally, the proposed approach has proven to perform better than other algorithms.