Rosyadi, Fuad Dary (2019) Rancang Bangun Mekanisme Computation Offloading Berbasis Model DMIPS untuk Peningkatan Kinerja Perangkat IoT Dalam Pemrosesan Data Sensor Pada Lingkungan Terdistribusi. Other thesis, Institut Teknologi Sepuluh Nopember.
|
Text
05111540000089-Undergraduate_Theses.pdf Restricted to Repository staff only Download (1MB) | Request a copy |
Abstract
Computation Offloading memberikan solusi dari permasalahan akan keterbatasan sumber daya pada jaringan sensor terdistribusi. Namun pada penerapannya Computation Offloading tidak selalu menguntungkan, dan bahkan dapat merugikan dikarenakan kondisi lingkungan yang cenderung dinamis dan tidak stabil. Sehingga Computation Offloading membutuhkan decision making yang tepat sehingga pemrosesan dapat dilakukan secara optimal. Untuk mengetahui pengaruh mekanisme computation offloading pada perangkat pada jaringan sensor terdistribusi, digunakan model pengukuran beban kerja berdasarkan DMIPS (Dhrystone Millions of Instruction per Second). Ujicoba dilakukan dengan membandingkan energi dan waktu eksekusi yang digunakan pada perangkat yang menerapkan mekanisme offloading menggunakan model DMIPS dengan yang hanya dikerjakan secara lokal dan offloading pada cloud server saja. Berdasarkan hasil ujicoba skenario yang dilakukan, computation offloading dengan model DMIPS meningkatkan kinerja proses komputasi dengan hasil waktu eksekusi 22.246 detik lebih cepat dari eksekusi skenario secara lokal dan 9.15 detik lebih cepat dari eksekusi skenario secara offloading serta konsumsi energi 82.139 joule lebih efisien dari eksekusi skenario secara lokal, dan 17.912 joule lebih efisien dari eksekusi skenario secara offloading.
=====================================================================================================================================
Computation Offloading provides a solution to the problem of limited resources in distributed sensor networks. However, in its application, Computation Offloading is not always profitable, and can even be detrimental due to environmental conditions that tend to be dynamic and unstable. So, Computation Offloading requires the right decision making so that processing can be done optimally. To determine the effect of the computation offloading mechanism on distributed sensor network devices, a workload measurement model is used based on DMIPS (Dhrystone Millions of Instruction per Second). The test is carried out by comparing the energy and execution time used on devices that apply the offloading mechanism using the DMIPS model with only being done locally and offloading the cloud server only. Based on the results of the test scenarios performed, computing offloading with the DMIPS model improves computational process performance with results of execution time of 22,246 seconds faster than execution of local scenarios and 9.15 seconds faster than execution of offloading scenarios and energy consumption of 82,139 joules more efficiently than scenario execution local, and 17,912 joules are more efficient than offloading scenario execution.
| Item Type: | Thesis (Other) |
|---|---|
| Additional Information: | RSIf 004.678 Ros r-1 2019 |
| Uncontrolled Keywords: | Computation Offloading, DMIPS, Jaringan Sensor Terdistribusi, Raspberry Pi |
| Subjects: | T Technology > T Technology (General) > T58.5 Information technology. IT--Auditing T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5103.2 Wireless communication systems. Two way wireless communication |
| Divisions: | Faculty of Information and Communication Technology > Informatics > 55201-(S1) Undergraduate Thesis |
| Depositing User: | Fuad Dary Rosyadi |
| Date Deposited: | 23 Jul 2026 06:56 |
| Last Modified: | 23 Jul 2026 06:56 |
| URI: | http://repository.its.ac.id/id/eprint/65043 |
Actions (login required)
![]() |
View Item |
