Penggunaan Ant-Lion Optimizer Algorithm (ALO) dan Elite Opposition Based Learning (EOBL) Sebagai Task Scheduler Dalam Cloud Environtment

Hakim, Fatih Rian Hibatul (2024) Penggunaan Ant-Lion Optimizer Algorithm (ALO) dan Elite Opposition Based Learning (EOBL) Sebagai Task Scheduler Dalam Cloud Environtment. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 5027201066-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 February 2026. Download (2MB) | Request a copy

Abstract

Komputasi awan adalah sebuah inovasi di mana pada zaman internet ini, pengguna bisa mengakses kekuatan komputasi komputer secara daring melalui internet tanpa harus memiliki resource seperti Hardware, Software, dan Platform. Semakin majunya zaman tentu semakin tinggi juga demand untuk penggunaan komputasi awan, mengetahui fakta tersebut ladang komputasi awan merupakan potensi sekaligus tantangan pada suatu aspek komputasi awan yang disebut sebagai Task Scheduling. mengeksplorasi penerapan algoritma Ant-Lion Optimizer (ALO) yang di tingkatkan dengan Elite Opposition Based-Learning (EOBL) untuk penjadwalan tugas di lingkungan cloud dengan menggunakan media Cloudsim sebagai perantara percobaan. Penelitian ini bertujuan untuk mengevaluasi penggabungkan ALO dengan Elite Opposition-Based Learning (EOBL) untuk penjadwalan tugas di lingkungan komputasi awan. Implementasi dilakukan menggunakan CloudSim dengan 54 virtual machine di 18 host yang dikelola oleh 6 datacenter. Hasil menunjukkan bahwa ALO yang di tingkatkan dengan EOBL memiliki performa yang lebih baik dalam menangani penjadwalan tugas di cloud dibandingkan dengan algoritma standar ALO dan Genetic Algorithm pada beberapa parameter seperti Makespan, Average Start Timer, Average Finish Time, Total Scheduling Length, Throughput, Resource Utilization dan imbalance degree. Sedangkan pada algoritma ALO sendiri paling unggul bila pada parameter Makespan dengan selisih 5,226, Total Wait Time dengan selisih 27.536.924ms, Throughput sebesar 0,0173 lebih besar, Resource Utilization 3,7% lebih besar, dan sedikit peningkatan Energy consumption. Adapun untuk dataset SDSC, Algoritma ALO unggul pada parameter Energy Consumption, Total Scheduling Length dan Total Wait Time Sedangkan pada algoritma ALO EOBL unggul pada parameter Throughput, Makespan, Resource Utilization. Bisa disimpulkan bahwa kolaborasi ALO dan EOBL akan mendapatkan hasil All-rounder di mana ALO hanya memberikan beberapa peningkatar pada parameter yang telah dijabarkan.
==================================================================================================================================
Cloud computing is an innovation where in this internet age, users can access computer computing power online via the internet without having to have resources such as hardware, software and platforms. As times progress, of course the demand for the use of cloud computing becomes higher. Knowing this fact, the cloud computing field is both a potential and a challenge in an aspect of cloud computing called Task Scheduling. explores the application of the Ant-Lion Optimizer (ALO) algorithm enhanced with Elite Opposition Based-Learning (EOBL) for task scheduling in a cloud environment using Cloudsim media as an experimental intermediary. This research aims to evaluate combining ALO with Elite Opposition-Based Learning (EOBL) for task scheduling in a cloud computing environment. Implementation was carried out using CloudSim with 54 virtual machines on 18 hosts managed by 6 data centers. The results show that ALO enhanced with EOBL has better performance in handling task scheduling in the cloud compared to the standard ALO algorithm and Genetic Algorithm on several parameters such as Makespan, Average Start Timer, Average Finish Time, Total Scheduling Length, Throughput, Resource Utilization and imbalance degree. Meanwhile, the ALO algorithm itself is superior in the Makespan parameters with a difference of 5.226, Total Wait Time with a difference of 27,536,924ms, Throughput 0.0173 greater, Resource Utilization 3.7% greater, and a slight increase in Energy consumption. As for the SDSC dataset, the ALO algorithm is superior in the parameters Energy Consumption, Total Scheduling Length and Total Wait Time. Meanwhile, the ALO EOBL algorithm is superior in the parameters Throughput, Makespan, Resource Utilization. It can be concluded that the collaboration between ALO and EOBL will get All-rounder results where ALO only provides several improvements to the parameters that have been described.

Item Type:	Thesis (Masters)
Additional Information:	RSTI 006.78 FAT p 2023
Uncontrolled Keywords:	Komputasi awan, Task Scheduling, Ant-lion Optimizer Algorithm, Elite Opposition-Based Learning, Cloud Computing, Task Scheduling, Ant-lion Optimizer Algorithm, Elite Opposition-Based Learning
Subjects:	Q Science > QA Mathematics > QA76.585 Cloud computing. Mobile computing.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology
Depositing User:	Fatih Rian Hibatul Hakim
Date Deposited:	06 Feb 2024 02:15
Last Modified:	06 Nov 2024 08:59
URI:	http://repository.its.ac.id/id/eprint/106233

Actions (login required)

View Item