Priority and dynamic quantum time algorithms for central processing unit scheduling

• Main Author: Mohammed, Maysoon A.
• Format: Thesis
• Language: English
• Published: 2018
• Subjects: QA75 Electronic computers. Computer science
• Online Access: http://umpir.ump.edu.my/id/eprint/23477/1/Priority%20and%20dynamic%20quantum%20time%20algorithms%20for%20central%20processing%20unit%20scheduling.pdf; http://umpir.ump.edu.my/id/eprint/23477/

Central Processing Unit scheduling is a key concept in computer multitasking, multiprocessing operating system and real‐ time operating system designs. Scheduling refers to the way processes are assigned to run on the Central Processing Unit. Central Processing Unit is scheduled using different types of scheduling algorithms. One of the most widely algorithm used in scheduling with sharing and batch operating systems is Round Robin. Round Robin algorithm arrange and choose all elements in a group equally in time. In Round Robin scheduling, processes get fair share of quantum time of the Central Processing Unit. The quantum time is fixed, so, it is inversely proportional to context switches which leads to high average waiting time and high average turnaround time which degrades the overall performance of the system. Researchers proposed different formulas and algorithms to improve the quantum time of Round Robin scheduling in different ways considering dynamic time, setting priorities and setting different number of queues. However, limitations have been found with the consideration of these algorithms. In this thesis, a new formula has been imbedded in two algorithms to improve existing Round Robin algorithms in term of context switches, turnaround time and waiting time. These formula and algorithms consider the priority, dynamic quantum time and different number of queues in order to increase the throughput of the Central Processing Unit. The proposed formula has been validated and analysed using excel software in order to access the new formula and compare with other existing formulas. In excel, a graph has been produced to show the minimum and maximum values of the new dynamic quantum time after setting the parameters of the proposed formula (priority, burst time and the current quantum time). The proposed algorithms (Priority Dynamic Quantum Time and Multi Priority Dynamic Quantum Time Algorithms) are implemented using JAVA programming language and validated using Key Performance Indicators equations. A comparison of the proposed algorithms with existing algorithms of Round Robin is presented to emphasize the importance of the proposed algorithms in order to improve the performance of the scheduling algorithm of the Central Processing Unit. After the experiments, the proposed algorithms successfully improved the throughput of the Central Processing Unit (i.e. the number of completed processes in each round is increased, number of context switches, average of turnaround and average of waiting time is decreased). For statistical analysis, a Wilcoxon non-parametric test is used to test and validate the results of the experiments. This study illustrated some contributions, 1. A new formula has been derived containing all components of the processes, burst time, quantum time and priority and 2. A new algorithms with different numbers of queues to improve the performance of the Central Processing Unit. The combination of dynamic quantum time and different priorities in single and multi-queue of Round Robin has been shown a significant contribution to increase the throughput of the Central Processing Unit.