An Adaptive Deep Learning Framework with Enhanced Attention for Precise Load Forecasting in Cloud Computing Environments
Received: 10 January 2026 | Revised: 31 January 2026, 7 February 2026, and 15 February 2026 | Accepted: 16 February 2026 | Online: 4 April 2026
Corresponding author: Kusuma Nidadavolu
Abstract
With the rapid growth of cloud computing services, accurate load forecasting and balancing have become critical for efficient resource utilization and system responsiveness. This paper proposes an adaptive deep learning model that integrates Convolutional Neural Networks (CNN), Bidirectional Long Short-Term Memory (Bi-LSTM), and an attention mechanism for precise multi-step load prediction. The CNN extracts local spatial features, the Bi-LSTM captures bidirectional temporal dependencies, and the attention mechanism dynamically focuses on the most relevant input segments, improving sensitivity to sudden load variations and anomalies. Evaluated on multi-step horizons (1–15 steps ahead) using a real-world IoT gateway and the Google Cluster Trace datasets, the model achieves average values of MSE at 0.0023, MAE at 0.035, and R² at 0.92 across horizons. These results indicate superior accuracy and explained variance compared to state-of-the-art baselines, making the approach highly suitable for cloud providers aiming to optimize resource provisioning and reduce operational costs.
Keywords:
attention mechanism, convolutional neural networks, cloud computing, deep learning, resource provision, time-series dataDownloads
References
R. Atat, L. Liu, J. Wu, G. Li, C. Ye, and Y. Yang, ''Big Data Meet Cyber-Physical Systems: A Panoramic Survey,'' IEEE Access, vol. 6, pp. 73603–73636, 2018. DOI: https://doi.org/10.1109/ACCESS.2018.2878681
A. Perçuku, D. Minkovska, and N. Hinov, ''Enhancing Electricity Load Forecasting with Machine Learning and Deep Learning,'' Technologies, vol. 13, no. 2, Feb. 2025, Art. no. 59. DOI: https://doi.org/10.3390/technologies13020059
D. A. Shafiq, N. Z. Jhanjhi, and A. Abdullah, ''Load balancing techniques in cloud computing environment: A review,'' Journal of King Saud University - Computer and Information Sciences, vol. 34, no. 7, pp. 3910–3933, July 2022. DOI: https://doi.org/10.1016/j.jksuci.2021.02.007
P. Vyas, K. M. Ragothaman, A. Chauhan, and B. Rimal, ''Data augmentation and generative machine learning on the cloud platform,'' International Journal of Information Technology, vol. 16, no. 8, pp. 4833–4843, Dec. 2024. DOI: https://doi.org/10.1007/s41870-024-02104-5
Y. Himeur, A. N. Sayed, A. Alsalemi, F. Bensaali, and A. Amira, ''Edge AI for Internet of Energy: Challenges and perspectives,'' Internet of Things, vol. 25, Apr. 2024, Art. no. 101035. DOI: https://doi.org/10.1016/j.iot.2023.101035
P. W. Tien, S. Wei, J. Darkwa, C. Wood, and J. K. Calautit, ''Machine Learning and Deep Learning Methods for Enhancing Building Energy Efficiency and Indoor Environmental Quality – A Review,'' Energy and AI, vol. 10, Nov. 2022, Art. no. 100198. DOI: https://doi.org/10.1016/j.egyai.2022.100198
S. Verma and A. Bala, ''ETSA-LP: Ensemble Time-Series Approach for Load Prediction in Cloud,'' Computing and Informatics, vol. 43, no. 1, pp. 64–93, 2024. DOI: https://doi.org/10.31577/cai_2024_1_64
H. Zhang, J. Li, and H. Yang, ''Cloud computing load prediction method based on CNN-BiLSTM model under low-carbon background,'' Scientific Reports, vol. 14, no. 1, Aug. 2024, Art. no. 18004. DOI: https://doi.org/10.1038/s41598-024-68339-1
J. Wang, J. Pan, F. Esposito, P. Calyam, Z. Yang, and P. Mohapatra, ''Edge Cloud Offloading Algorithms: Issues, Methods, and Perspectives,'' ACM Computing Surveys, vol. 52, no. 1, pp. 1–23, Jan. 2020. DOI: https://doi.org/10.1145/3284387
M. I. Khaleel, “A dynamic weight–assignment load balancing approach for workflow scheduling in edge-cloud computing using ameliorated moth flame and rock hyrax optimization algorithms,” Future Generation Computer Systems, vol. 155, pp. 465–485, June 2024. DOI: https://doi.org/10.1016/j.future.2024.02.025
N. Hogade and S. Pasricha, ''A Survey on Machine Learning for Geo-Distributed Cloud Data Center Management,'' IEEE Transactions on Sustainable Computing, vol. 8, no. 1, pp. 15–31, Jan. 2023. DOI: https://doi.org/10.1109/TSUSC.2022.3208781
S. Verma and A. Bala, ''Efficient Auto‐scaling for Host Load Prediction through VM migration in Cloud,'' Concurrency and Computation: Practice and Experience, vol. 36, no. 4, Feb. 2024, Art. no. e7925. DOI: https://doi.org/10.1002/cpe.7925
S. Simaiya et al., ''A hybrid cloud load balancing and host utilization prediction method using deep learning and optimization techniques,'' Scientific Reports, vol. 14, no. 1, Jan. 2024, Art. no. 1337. DOI: https://doi.org/10.1038/s41598-024-51466-0
M. Valizadeh and S. J. Wolff, ''Convolutional Neural Network applications in additive manufacturing: A review,'' Advances in Industrial and Manufacturing Engineering, vol. 4, May 2022, Art. no. 100072. DOI: https://doi.org/10.1016/j.aime.2022.100072
A. Goel, A. K. Goel, and A. Kumar, ''The role of artificial neural network and machine learning in utilizing spatial information,'' Spatial Information Research, vol. 31, no. 3, pp. 275–285, June 2023. DOI: https://doi.org/10.1007/s41324-022-00494-x
S. Zairi and M. Freihat, ''Electric Load Forecasting using Machine Learning for Peak Demand Management in Smart Grids,'' Engineering, Technology & Applied Science Research, vol. 15, no. 3, pp. 23335–23346, June 2025. DOI: https://doi.org/10.48084/etasr.10687
H. Toumi, Z. Brahmi, and M. M. Gammoudi, ''RTSLPS: Real time server load prediction system for the ever-changing cloud computing environment,'' Journal of King Saud University - Computer and Information Sciences, vol. 34, no. 2, pp. 342–353, Feb. 2022. DOI: https://doi.org/10.1016/j.jksuci.2019.12.004
"google/cluster-data." Google, [Online]. Available: https://github.com/google/cluster-data.
Downloads
How to Cite
License
Copyright (c) 2026 Kusuma Nidadavolu, Somasekhar Giddaluru
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain the copyright and grant the journal the right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) after its publication in ETASR with an acknowledgement of its initial publication in this journal.
