Evidence record · UDINUS · SDGs
- Unit
- Rektorat UDINUS
- Period
- 2021
- Goals
- Record
- 6b6e1d5b-20dc-45f7-bcd6-425d5c9ee9fe
- Published
- 2026-10-30
UDINUS conducted research on improving the Virtual Machine Selection mechanism in dynamic virtualization consolidation to achieve robust energy-efficient consumption. The activity involved Shidik, Guruh Fajar; Nugraha, Adhitya; Andono, Pulung Nurtantio; Noersasongko, Edi; and Kusuma, Edi Jaya. The research focused on Virtual Machine Selection and examined unsupervised methods combined with a rule-based MAX function.
The proposed methods were K-Means and Fuzzy C-Mean (FCM), each combined with the rule-based MAX function. Both proposed methods were simulated, compared, and evaluated using a public workload dataset. The evaluation measured energy consumption in kilowatt-hour using the CloudSim Environment. The proposed methods were compared with statistical RC and MMT VM selection methods. The simulation result showed that the highest decrement of energy consumption was achieved by K-Means rule-based in five clusters, reaching 16.51 kWh. FCM rule-based followed, reaching 15.7 kWh.
The research also used the statistical Friedman Test to evaluate energy-efficiency improvement. The test produced a p-value of 0.0105 and indicated a significant difference in energy consumption among all tested methods. K-Means Rule-Based achieved the first rank in the Friedman Test. Through this activity, UDINUS investigated energy-efficient consumption by comparing Virtual Machine Selection methods in a simulated environment. The evidence concerns the methods, dataset, measurement environment, comparison methods, and reported findings described in the source.