Optimizing the fog service placement with r3gp: a rotation-guided greedy genetic particle algorithm
The Fog Computing paradigm emerged as a complementary solution to the Cloud Computing to bring application processing to edge computing devices, which interconnect with typical Internet of Things (IoT) devices. However, the limited capacity of edge nodes poses some challenges in managing the resourc...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Estado: | Versión publicada |
| Fecha de publicación: | 2023 |
| País: | Brasil |
| Institución: | Universidade Federal de Sergipe (UFS) |
| Repositorio: | Repositório Institucional da UFS |
| Idioma: | portugués |
| OAI Identifier: | oai:oai:ri.ufs.br:repo_01:riufs/19477 |
| Acceso en línea: | https://ri.ufs.br/jspui/handle/riufs/19477 |
| Access Level: | acceso abierto |
| Palabra clave: | Internet das coisas Computação em nuvem Service placement Fog computing Cloud computing Edge computing Optimization Internet of things CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO |
| Sumario: | The Fog Computing paradigm emerged as a complementary solution to the Cloud Computing to bring application processing to edge computing devices, which interconnect with typical Internet of Things (IoT) devices. However, the limited capacity of edge nodes poses some challenges in managing the resources available to distributed applications. Service placement in Fog Computing is an NP-complete problem that consists of managing the decision on which Fog node the service of an IoT application will run. If there is not enough resource in the Fog, the application is sent to the Cloud. This work consists of optimizing the Fog Service Placement Problem for the execution of IoT applications, applying a case study regarding vehicle collisions on urban roads. The problem is formulated as a Constraint Satisfaction Problem for optimization of five objective functions: makespan, energy consumption gap, CPU load-balancing, memory load-balancing and bandwidth load-balancing. In this work, an algorithm for optimization of the problem, named Rotation-Guided Greedy Genetic Particle (R3GP), is proposed. The study is conducted with an in silico experiment that compares the algorithm with others found in the literature. Statistical results show that R3GP can outperform the compared algorithms, mainly in optimizing the energy consumption gap metric. |
|---|