Λεπτομέρειες βιβλιογραφικής εγγραφής
| Τίτλος: |
Frequency‐Domain‐Enhanced Dynamic Graph Neural Network for Cloud Cluster Workload Prediction. |
| Συγγραφείς: |
Wu, Hao, Qian, Jiahao, Li, Haotong, Chen, Qidong, Murray, Richard |
| Πηγή: |
International Journal of Intelligent Systems; 7/23/2026, Vol. 2026, p1-14, 14p |
| Θεματικοί όροι: |
Graph neural networks, Frequency-domain analysis, Cloud computing, Scheduling, Computer performance |
| Περίληψη: |
Accurate workload prediction is fundamental to enhancing scheduling efficiency, optimizing resource utilization, and ensuring system stability within large‐scale cloud cluster environments. However, the high concurrency of multitenant tasks introduces significant instability, high‐frequency fluctuations, and unpredictable burstiness in resource demands, posing substantial challenges to existing scheduling and management. To address these issues, we propose in this paper a frequency‐domain‐enhanced dynamic graph neural network named FEDG, which integrates a dynamic graph structure learning mechanism with frequency‐domain convolution so that evolving internode dependencies and periodic characteristics of workload sequences can be modeled in a unified framework. Comprehensive experiments conducted on a real‐world workload dataset derived from the Google Cluster trace demonstrate that FEDG achieves an MSE of 0.0998, MAE of 0.1313, and R2 score of 0.9233. The results indicate that FEDG provides an effective solution for predicting large‐scale cluster workloads and could offer useful support for resource scheduling and system management in cloud environments. [ABSTRACT FROM AUTHOR] |
|
Copyright of International Journal of Intelligent Systems is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) |
| Βάση Δεδομένων: |
Complementary Index |