Autonomous scheduling mechanism based on energy awareness for improving resource allocation in serverless IoT edge
Abstract The energy-aware scheduling mechanism in serverless computing enables systems to be allocated to IoT devices connected to the network’s edge efficiently based on the energy status of active nodes. In this approach, resource allocation is performed in real-time and according to the energy ch...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-07-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-04214-x |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract The energy-aware scheduling mechanism in serverless computing enables systems to be allocated to IoT devices connected to the network’s edge efficiently based on the energy status of active nodes. In this approach, resource allocation is performed in real-time and according to the energy changes of the nodes, the complexity of which arises from the direct dependence of the energy level on the capacity to respond to requests. This choice is complex since allocating the necessary resources to process requests depends on the energy available in the active nodes. Therefore, to optimize resource allocation and increase access time, selecting and executing pre-schedulers is necessary based on the prediction of the energy level of the active nodes. In this article, we introduce a scheduler selection mechanism called an autonomous energy-aware scheduler, whose design is based on the energy position of the active nodes in the network. In addition, a mechanism for improving system uptime is proposed to increase the duration of the energy reduction of active nodes. The efficiency of the proposed approach was evaluated utilizing three separate load distribution patterns (exponential, Poisson, and exponential-Poisson), and the results indicate the prevention of energy waste and an average reduction of 1.66% in energy consumption. Also, the network uptime is improved by 8.6% compared to other methods. In addition, the proposed method has maintained performance continuity while ensuring failure resistance in all situations. The results demonstrate the high efficiency of the proposed approach in optimizing energy consumption and enhancing the resilience and stability of serverless systems. |
|---|---|
| ISSN: | 2045-2322 |