SalisuIbrahimYusuf-Grouped-Half-Life

A Grouped Half-Life Quantum
Time Round Robin Scheduling
(GHLVQTRR) Algorithm for CPU
Process
YUSUF Salisu Ibrahim, S E Abdullahi PhD
Computer Science Department, BAZE University, Nigeria.
Computer Science Department, Nile University of Nigeria, Nigeria.
3rd Big Data Analytics and Innovation Conference, 22-25 November, NDC, Abuja, Nigeria
Operating System and CPU process
Scheduling
• Intermediary between the computer resources and
computer users.
• CPU is the most scarce computer resource.
• CPU Scheduling algorithms are used to coordinate the
processing activities.
3rd Big Data Analytics and Innovation Conference, 22-25 November, NDC, Abuja, Nigeria
CPU Scheduling Algorithms
1. First-In-First-Out (FIFO)/First-Come-First-Serve
(FCFS).
2. Shortest-Job-First (SJF).
3. Longest-Job-First (LJF).
4. Priority Based Scheduling.
5. Round Robin (RR).
3rd Big Data Analytics and Innovation Conference, 22-25 November, NDC, Abuja, Nigeria
Evaluation Criteria
• CPU Utilization.
• Throughput.
• Turnaround Time.
• Waiting Time.
• Response Time.
• Context Switching.
3rd Big Data Analytics and Innovation Conference, 22-25 November, NDC, Abuja, Nigeria
Round Robin Algorithm
• Processes take turns based of a TIME SLICE aka
QUANTUM TIME. Traditionally uses the average of
the processes burst times.
• Very effective in a multiprogramming environment.
Challenges in determining QT
• Small QT  large context switching.
• Large QT  system behaves like FCFS
3rd Big Data Analytics and Innovation Conference, 22-25 November, NDC, Abuja, Nigeria
Variations Round Robin Algorithm
• Dynamic Round Robin with controlled pre-emption
(DRRCP) (Ashiru, et al., 2014).
• Priority based Round Robin CPU scheduling algorithms
for real time system (Singh & Deepa, 2012).
• Variable Quantum Time (VQT) algorithm which is based
on averaging technique (Yashasvini, 2013).
• In Even Odd Round Robin (EORR), there are two QT
(QT1 and QT2).
• …
3rd Big Data Analytics and Innovation Conference, 22-25 November, NDC, Abuja, Nigeria
Variations Round Robin Algorithm
• Average Mid Max Round Robin (AMMRR),
• Ascending Quantum Minimum and Maximum Round
Robin (AQMMRR). (Dawood, 2012).
• The Half Life Variable Quantum Time Round Robin
(HLVQTRR) (Ashiru, et al., 2014).
▫ Takes ½ of the individual burst time of process as QT for the
process.
▫ Limitation observed: relatively very large processes will spent
very long time executing
3rd Big Data Analytics and Innovation Conference, 22-25 November, NDC, Abuja, Nigeria
A Grouped HLVQTRR
• An optimization of HLVQTRR
• Recursively grouped group processes based on their
burst times.
• ½ of the average of the processes is made to be the
QT of process less than or equal to the half of the
average.
3rd Big Data Analytics and Innovation Conference, 22-25 November, NDC, Abuja, Nigeria
Comparing G-HLVQTRR and HLVQTRR
Case 1: 7 processes, range: 84 – 93 average: 88
Criteria
HLVQTRR
G-HLVQTRR
Average Waiting Time
398.29
398.86
Number of context
Switching.
7
7
Average Turn Around
Time
486.71
487.29
3rd Big Data Analytics and Innovation Conference, 22-25 November, NDC, Abuja, Nigeria
Comparing G-HLVQTRR and HLVQTRR
Case 2: 20 processes, range: 3 – 21 average: 9.4
Criteria
HLVQTRR
G-HLVQTRR
Average Waiting Time
127.33
108.15
Number of context
Switching.
20
15
Average Turn Around
Time
133.53
127
3rd Big Data Analytics and Innovation Conference, 22-25 November, NDC, Abuja, Nigeria
Discussion
• When the range of the process difference is wide.
HLVQTRR algorithm was slight better
• With a wider range the proposed algorithm
much better
Other Applications of Scheduling
algorithm.
• Parallel computing: task sharing.
• Military: Optimal use of scarce resource.
• Personal: time management.