QoS-aware CR-BM-based hybrid framework to improve the fault tolerance of fog devices
With the evolution of the Internet of Things (IoT), the use of smart devices has completely changed the day-to-day life of the human being. IoT devices are of flexible use which is implemented to sense the environment and data efficiently. However, these devices have some constrained capabilities co...
| Authors: | , |
|---|---|
| Format: | article |
| Status: | Published version |
| Publication Date: | 2021 |
| Country: | México |
| Institution: | UNIVERSIDAD NACIONAL AUTÓNOMA DE MÉXICO |
| Repository: | Journal of Applied Research and Technology |
| Language: | English |
| OAI Identifier: | oai:ojs2.localhost:article/1493 |
| Online Access: | https://jart.icat.unam.mx/index.php/jart/article/view/1493 |
| Access Level: | Open access |
| Keyword: | IoT QoS fault tolerance checkpoint replication bee-mutation service placement |
| Summary: | With the evolution of the Internet of Things (IoT), the use of smart devices has completely changed the day-to-day life of the human being. IoT devices are of flexible use which is implemented to sense the environment and data efficiently. However, these devices have some constrained capabilities concerning fault tolerance, computation cost, and storage. This requires an improved framework and algorithms for performing effective operations. In this paper, a hybrid framework is proposed, which incorporates the various IoT devices in fog environments to enhance fault tolerance. The proposed framework implements a novel QoS-aware technique based on the combination of checkpoints and replication (CR) for diagnosing faults and the bee-mutation (BM) algorithm for optimal placement of service. A fog service monitor is established to observe the performance of fog nodes. Both the CR module and BM module access the service monitor to ensure that the proposed hybrid framework is fault-tolerant. Furthermore, the proposed CR-BM-based hybrid framework has been evaluated for its performance by using various performance metrics. In the comparative analysis, it is observed that the proposed hybrid framework outperforms the existing genetic algorithm-based framework. |
|---|