Abstract
Cloud computing has recently gained popularity due to its cost-effective and high-quality services. Cloud-native systems are expected to host more than 95% of digital workloads. Cloud service providers face two significant challenges: real-time workload predictions and effective resource management. Furthermore, allocating resources over time may result in a suboptimal execution environment due to considerable increases and decreases in workload that follow time-dependent patterns. Recent breakthroughs in deep learning have garnered widespread favor for predicting extremely nonlinear cloud workloads; nevertheless, they have been unable to generalize inter cluster workload forecasting due to inadequate workload data at the beginning of each cluster. Furthermore, the distribution disparity across distinct cluster workloads is caused by a variety of elements, making it difficult to reuse current data or models directly. To overcome these challenges, we propose ADAPT, which relies on Attention-Driven Domain Adaptation. First, we use LSTM architecture as the backbone of our model. Moreover, we construct a strategically shared attention module to transmit relevant knowledge from the source domain to the target domain by inducing domain-invariant latent features and retraining domain-specific features. Lastly, adversarial training is used to increase the model’s resilience and predictive accuracy. Comprehensive experimental evaluations indicate that our proposed approach significantly outperforms existing baselines.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Aswolinskiy, W., Hammer, B.: Unsupervised transfer learning for time series via self-predictive modelling-first results. In: Proceedings of the Workshop on New Challenges in Neural Computation (NC2), vol. 3 (2017)
Caglar, F., Gokhale, A.: ioverbook: intelligent resource-overbooking to support soft real-time applications in the cloud. In: 2014 IEEE 7th International Conference on Cloud Computing, pp. 538–545. IEEE (2014)
da Costa, P.R.d.O., Akçay, A., Zhang, Y., Kaymak, U.: Remaining useful lifetime prediction via deep domain adaptation. Reliability Eng. Syst. Safety 195, 106682 (2020)
Gong, Z., Gu, X., Wilkes, J.: Press: predictive elastic resource scaling for cloud systems. In: 2010 International Conference on Network and Service Management, pp. 9–16. IEEE (2010)
Goodfellow, I., et al.: Generative adversarial networks. Commun. ACM 63(11), 139–144 (2020)
Gretton, A., Borgwardt, K., Rasch, M., Schölkopf, B., Smola, A.: A kernel method for the two-sample-problem. Adv. Neural Inform. Process. Syst. 19 (2006)
Hershey, J.R., Olsen, P.A.: Approximating the kullback leibler divergence between gaussian mixture models. In: 2007 IEEE International Conference on Acoustics, Speech and Signal Processing-ICASSP 2007, vol. 4, pp. IV–317. IEEE (2007)
Hoffman, J., et al.: Cycada: cycle-consistent adversarial domain adaptation. In: International Conference on Machine Learning, pp. 1989–1998. PMLR (2018)
Hu, H., Tang, M., Bai, C.: Datsing: data augmented time series forecasting with adversarial domain adaptation. In: Proceedings of the 29th ACM International Conference on Information & Knowledge Management, pp. 2061–2064 (2020)
Jiang, J., Zhai, C.: Instance weighting for domain adaptation in nlp. ACL (2007)
Kim, Y.M., Song, S., Koo, B.M., Son, J., Lee, Y., Baek, J.G.: Enhancing long-term cloud workload forecasting framework: Anomaly handling and ensemble learning in multivariate time series. IEEE Trans. Cloud Comput. (2024)
Leka, H.L., Fengli, Z., Kenea, A.T., Tegene, A.T., Atandoh, P., Hundera, N.W.: A hybrid cnn-lstm model for virtual machine workload forecasting in cloud data center. In: 2021 18th International Computer Conference on Wavelet Active Media Technology and Information Processing (ICCWAMTIP), pp. 474–478. IEEE (2021)
Long, M., Cao, Z., Wang, J., Jordan, M.I.: Conditional adversarial domain adaptation. Adv. Neural Inform. Process. Syst. 31 (2018)
Patel, Y.S., Bedi, J.: Mag-d: a multivariate attention network based approach for cloud workload forecasting. Futur. Gener. Comput. Syst. 142, 376–392 (2023)
Podolskiy, V., Jindal, A., Gerndt, M., Oleynik, Y.: Forecasting models for self-adaptive cloud applications: a comparative study. In: 2018 ieee 12th international conference on self-adaptive and self-organizing systems (SASO), pp. 40–49. IEEE (2018)
Roy, N., Dubey, A., Gokhale, A.: Efficient autoscaling in the cloud using predictive models for workload forecasting. In: 2011 IEEE 4th International Conference on Cloud Computing, pp. 500–507. IEEE (2011)
Saxena, D., Kumar, J., Singh, A.K., Schmid, S.: Performance analysis of machine learning centered workload prediction models for cloud. IEEE Trans. Parallel Distrib. Syst. 34(4), 1313–1330 (2023)
Sherstinsky, A.: Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network. Physica D 404, 132306 (2020)
Subramanian, S., Kannammal, A.: Real time non-linear cloud workload forecasting using the holt-winter model. In: 2019 10th International Conference on Computing, Communication and Networking Technologies (ICCCNT), pp. 1–6. IEEE (2019)
Sun, B., Feng, J., Saenko, K.: Return of frustratingly easy domain adaptation. In: Proceedings of the AAAI Conference On Artificial Intelligence, vol. 30 (2016)
Tirmazi, M., et al.: Borg: the next generation. In: Proceedings of the Fifteenth European Conference on Computer Systems, pp. 1–14 (2020). https://doi.org/10.1145/3342195.338751
Valarmathi, K., Kanaga Suba Raja, S.: Resource utilization prediction technique in cloud using knowledge based ensemble random forest with lstm model. Concurrent Eng. 29(4), 396–404 (2021)
Wang, M., Deng, W.: Deep visual domain adaptation: a survey. Neurocomputing 312, 135–153 (2018)
Wu, Y., Liu, J., Wang, C., Xie, X., Shi, G.: Graph transformer and LSTM attention for VNF multi-step workload prediction in sfc. IEEE Trans. Netw. Service Manag. (2024)
Xi, H., Yan, C., Li, H., Xiao, Y.: An attention-based recurrent neural network for resource usage prediction in cloud data center. J. Phys. Conf. Ser. 2006, 012007 (2021)
Yazdanian, P., Sharifian, S.: E2lg: a multiscale ensemble of lstm/gan deep learning architecture for multistep-ahead cloud workload prediction. J. Supercomput. 77, 11052–11082 (2021)
Zhang, Q., Zhani, M.F., Zhang, S., Zhu, Q., Boutaba, R., Hellerstein, J.L.: Dynamic energy-aware capacity provisioning for cloud computing environments. In: Proceedings of the 9th International Conference on Autonomic Computing, pp. 145–154 (2012)
Zheng, H., et al.: Energy optimisation in cloud datacentres with mc-tide: mixed channel time-series dense encoder for workload forecasting. Appl. Energy 374, 123903 (2024)
Acknowledgments
This research was partly supported by the 1) Korea Institute of Science and Technology Information(KISTI) 2) Institute of Information & communications Technology Planning & Evaluation (IITP) grant funded by the Korea government(MSIT) (No. RS-2023-00220631, Edge Cloud Reference Architecture Standardization for Low Latency and Lightweight Cloud Service).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2025 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Mahbub, N.I., Sinthia, A.K., Jeon, M., Park, J., Huh, EN. (2025). ADAPT: Attention-Driven Domain Adaptation for Inter-cluster Workload Forecasting in Cloud Data Centers. In: Wang, Y., Zhang, LJ. (eds) CLOUD Computing – CLOUD 2024. CLOUD 2024. Lecture Notes in Computer Science, vol 15423. Springer, Cham. https://doi.org/10.1007/978-3-031-77153-8_6
Download citation
DOI: https://doi.org/10.1007/978-3-031-77153-8_6
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-77152-1
Online ISBN: 978-3-031-77153-8
eBook Packages: Computer ScienceComputer Science (R0)