Cloud Task Scheduling Menggunakan Artificial Fish Swarm Algorithm dan Opposition Based Learning

Simanjuntak, Jonathan Leonardo Hasiholan (2024) Cloud Task Scheduling Menggunakan Artificial Fish Swarm Algorithm dan Opposition Based Learning. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 05111940000150-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2026. Download (4MB) | Request a copy

Abstract

Saat ini, cloud computing adalah salah satu teknologi yang berkembang dengan pesat, dan perlahan – lahan menarik perhatian banyak orang. Seiring dengan meningkatnya kebutuhan akan sumber daya cloud, cloud service provider terus menambah jumlah data center yang dimilikinya agar dapat memenuhi kebutuhan pasar. Sehingga, diperlukan adanya suatu sistem manajemen sumber daya cloud yang baik bagi para cloud service provider agar sumber daya yang dimiliki dapat di kelola dengan lebih optimal tanpa melanggar service-level agreement (SLA) yang telah ditetapkan. Salah satu fokus utama dari sistem manajemen cloud adalah proses penjadwalan tugas. Penjadwalan tugas yang dilakukan akan mempengaruhi tingkat efisiensi pemanfataan sumber daya cloud secara keseluruhan. Oleh karena itu, dibutuhkan sebuah mekanisme penjadwalan tugas yang efisien di lingkungan cloud computing. Dengan demikian, penulis mengusulkan sebuah sistem penjadwalan tugas cloud dengan menggunakan Artificial Fish Swarm Algorithm (AFSA) dan Opposition Based Learning (OBL). Penelitian ini dilakukan dengan menjalankan dua skenario simulasi penjadwalan tugas cloud yang berbeda menggunakan CloudSim, sebuah alat simulasi lingkungan cloud. Pada skenario pertama, sistem penjadwalan tugas menggunakan Artificial Fish Swarm Algorithm (AFSA). Sedangkan pada skenario kedua, sistem penjadwalan tugas menggunakan Artificial Fish Swarm Algorithm yang digabungkan dengan Opposition Based Learning (AFSA - OBL). Hasil dari uji coba menunjukkan bahwa Artificial Fish Swarm Algorithm berhasil memberikan hasil optimasi penjadwalan tugas yang cenderung lebih baik daripada Genetic Algorithm. Perbaikan hasil optimasi yang diberikan diantaranya adalah menurunkan nilai makespan sebesar 28% hingga 45%, dan memberikan peningkatan pada throughput sebesar 34% hingga 82%. Namun, implementasi Opposition Based Learning yang dilakukan tidak memberikan dampak yang signifikan terhadap nilai hasil optimasi Artificial Fish Swarm Algorithm. Perbaikan nilai yang diberikan oleh Opposition Based Learning berada di kisaran 1,5% untuk nilai makespan, dan 1,39% untuk nilai througput.
=============================================================================================================================
Today, cloud computing is one of the fastest growing technologies, and is slowly attracting the attention of many people. Along with the increasing need for cloud resources, cloud service providers continue to increase the number of data centers they have in order to meet market needs. Thus, a good cloud resource management system is needed for cloud service providers so that their resources can be managed more optimally without violating the established service-level agreement. One of the main focuses of the cloud management system is the task scheduling process. The task scheduling process will affect the overall efficiency level of cloud resource utilization. Therefore, an efficient task scheduling mechanism is needed in a cloud computing environment. Thus, the author proposes a cloud task scheduling system using Artificial Fish Swarm Algorithm (AFSA) and Opposition Based Learning (OBL). This research was conducted by running two different cloud task scheduling simulation scenarios using CloudSim, a cloud environment simulation tool. In the first scenario, the task scheduling system uses the Artificial Fish Swarm Algorithm (AFSA). While in the second scenario, the task scheduling system uses Artificial Fish Swarm Algorithm combined with Opposition Based Learning (AFSA - OBL). The results show that the Artificial Fish Swarm Algorithm successfully provides task scheduling optimization results that tend to be better than Genetic Algorithm. Improvements in optimization results include reducing the makespan value by 28% to 45% and providing an increase in throughput by 34% to 82%. However, the implementation of Opposition Based Learning does not have a significant impact on the value of the Artificial Fish Swarm Algorithm optimization results. The value improvement provided by Opposition Based Learning is in the range of 1,5% for the makespan value, and 1.39% for the throughput value.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Artificial Fish Swarm Algorithm, Cloud Computing, CloudSim, Opposition Based Learning, Penjadwalan Tugas, Task Scheduling
Subjects:	Q Science > QA Mathematics > QA76.585 Cloud computing. Mobile computing.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Jonathan Leonardo Hasiholan Simanjuntak
Date Deposited:	11 Feb 2024 03:16
Last Modified:	11 Feb 2024 03:16
URI:	http://repository.its.ac.id/id/eprint/106718

Actions (login required)

View Item