A Chemistry-Based Optimization Algorithm for Quality of Service-Aware Multi-Cloud Service Compositions
The increasing complexity of cloud service composition demands innovative approaches that can efficiently optimize both functional requirements and quality of service (QoS) parameters. While several methods exist, they struggle to simultaneously minimize the number of combined clouds, examined servi...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-04-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/13/8/1351 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The increasing complexity of cloud service composition demands innovative approaches that can efficiently optimize both functional requirements and quality of service (QoS) parameters. While several methods exist, they struggle to simultaneously minimize the number of combined clouds, examined services, and execution time while maintaining a high QoS. This novelty of this paper is the chemistry-based approach (CA) that draws inspiration from the periodic table’s organizational principles and electron shell theory to systematically reduce the complexity associated with service composition. As chemical elements are organized in the periodic table and electrons organize themselves in atomic shells based on energy levels, the proposed approach organizes cloud services in hierarchical structures based on their cloud number, composition frequencies, cloud quality, and QoS levels. By mapping chemical principles to cloud service attributes—where service quality levels correspond to electron shells and service combinations mirror molecular bonds—an efficient framework for service composition is created that simultaneously addresses multiple objectives in QoS, NC, NEC, NES, and execution time. The experimental results demonstrated significant improvements over existing methods, such as Genetic Algorithms (GAs), Simulated Annealing (SA), and Tabu Search (TS), across multiple performance metrics, i.e., reductions of 14–33% are observed in combined clouds, while reductions of 20–85% are observed in examined clouds, and reductions of 74–98% are observed in examined services. Also, a reduction of 10–99% is observed in execution time, while fitness levels are enhanced by 1–14% compared to benchmarks. These results validate the proposed approach’s effectiveness in optimizing service composition while minimizing computational overhead in multi-cloud environments. |
|---|---|
| ISSN: | 2227-7390 |