Dynamic Energy-Aware Task Scheduling using Real-Time Resource Monitoring in Distributed Edge Environments
Fuente:
Zenodo
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Recurso digital |
| Veröffentlicht: |
Zenodo
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866901350374178816 |
|---|---|
| author | Vanapalli Jayanth Sai Chadalavada Venkata Sai Nitin Dr. Bharati Bidikar Vobilisetty Vandana Dedeepya Yelley Ankitha |
| author_facet | Vanapalli Jayanth Sai Chadalavada Venkata Sai Nitin Dr. Bharati Bidikar Vobilisetty Vandana Dedeepya Yelley Ankitha |
| contents | Edge computing environments support latency-sensitive applications but suffer from energy inefficiency and poor resource utilization due to dynamic workloads and heterogeneous nodes. Traditional scheduling methods such as Round Robin and heuristic-based approaches rely on static decisions, leading to increased energy consumption, higher latency, and uneven load distribution. This paper proposes a Dynamic Energy-Aware Task Scheduling Framework that combines real-time resource monitoring with a hybrid CNN-LSTM model to enable adaptive scheduling. The system continuously collects node-level telemetry (CPU, memory, temperature, energy) and predicts optimal node selection using a ranking-based approach refined by current load conditions. Experimental results show that the proposed method achieves 18–25% reduction in energy consumption, 20–30% lower latency, and 15–22% higher throughput, while maintaining balanced resource utilization, demonstrating its effectiveness for intelligent edge scheduling. |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_20025984 |
| institution | Zenodo |
| language | |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Dynamic Energy-Aware Task Scheduling using Real-Time Resource Monitoring in Distributed Edge Environments Vanapalli Jayanth Sai Chadalavada Venkata Sai Nitin Dr. Bharati Bidikar Vobilisetty Vandana Dedeepya Yelley Ankitha Edge Computing Task Scheduling Energy Awareness CNN-LSTM Real-Time Resource Monitoring Distributed Systems Deep Learning Edge computing environments support latency-sensitive applications but suffer from energy inefficiency and poor resource utilization due to dynamic workloads and heterogeneous nodes. Traditional scheduling methods such as Round Robin and heuristic-based approaches rely on static decisions, leading to increased energy consumption, higher latency, and uneven load distribution. This paper proposes a Dynamic Energy-Aware Task Scheduling Framework that combines real-time resource monitoring with a hybrid CNN-LSTM model to enable adaptive scheduling. The system continuously collects node-level telemetry (CPU, memory, temperature, energy) and predicts optimal node selection using a ranking-based approach refined by current load conditions. Experimental results show that the proposed method achieves 18–25% reduction in energy consumption, 20–30% lower latency, and 15–22% higher throughput, while maintaining balanced resource utilization, demonstrating its effectiveness for intelligent edge scheduling. |
| title | Dynamic Energy-Aware Task Scheduling using Real-Time Resource Monitoring in Distributed Edge Environments |
| topic | Edge Computing Task Scheduling Energy Awareness CNN-LSTM Real-Time Resource Monitoring Distributed Systems Deep Learning |
| url | https://doi.org/10.5281/zenodo.20025984 |