Beyond Reactive Scaling: A Review of AI-Driven Proactive and Context-Aware Auto-Scaling in Cloud-Edge Environments
DOI:
https://doi.org/10.15662/IJEETR.2024.0603009Keywords:
AI-driven auto-scaling, proactive scaling, context-aware computing, cloud-edge environments, machine learning, Kubernetes, workload prediction, edge computing, resource optimization, deep learningAbstract
The rapid nature of cloud-edge computing requires advanced resource scaling techniques. Though the popular reactive scaling is plagued by inefficiency and delay, this review presents AI-based proactive and context-aware auto-scaling methodologies that forecast workload changes and scale resources in advance. By reviewing the combination of machine learning and deep learning algorithms with cloud-native orchestration tools, this paper points out their scope in reducing latency, saving costs, and improving performance. In addition, we provide an overview of current methodologies, present context-aware decision-making mechanisms, and list open research problems. The aim of this paper is to direct future research into managing sustainable and intelligent cloud-edge infrastructures.
References
[1] Al-Dhuraibi, Y., Paraiso, F., Djarallah, N., & Merle, P. (2018). Elasticity in cloud computing: State of the art and research challenges. IEEE Transactions on Services Computing, 11(2), 430–447.
[2] Bermbach, D., Pröstl-Andrade, L., & Tai, S. (2020). Benchmarking auto-scaling strategies for cloud applications. Journal of Systems and Software, 168, 110638.
[3] Buyya, R., Srirama, S. N., Casale, G., & Varghese, B. (2021). Next-generation cloud computing: New trends and research directions. Future Generation Computer Systems, 115, 619–632.
[4] Chen, S., Wang, Z., Liu, W., & Zhang, Y. (2021). Federated learning for distributed auto-scaling in edge networks. IEEE Transactions on Services Computing, 14(6), 2105–2118.
[5] Chen, Y., Hu, Q., Liu, X., & Zhang, Y. (2020). Context-aware resource allocation in edge-cloud systems using Bayesian models. Computer Networks, 178, 107341.
[6] Dang, T. T., Truong, H. L., & Dustdar, S. (2019). AI-based resource auto-scaling for containerized applications. Proceedings of the IEEE International Conference on Cloud Computing, 192–199.
[7] Fang, Z., Wu, H., Ren, H., & Wang, G. (2023). Resource-aware deep reinforcement learning for edge-cloud orchestration. Journal of Parallel and Distributed Computing, 172, 1–12.
[8] Garcia Lopez, P., Montresor, A., Epema, D., Datta, A., Higashino, T., Iamnitchi, A., ... & Felber, P. (2022). AI-based predictive auto-scaling in cloud computing: A survey. Future Generation Computer Systems, 128, 205–220.
[9] Ghobaei-Arani, M., Souri, A., & Rahmanian, A. (2021). A hybrid metaheuristic algorithm for QoS-aware auto-scaling of cloud resources using container-based architecture. Concurrency and Computation: Practice and Experience, 33(6), e6067.
[10] Guo, Y., Zhang, Y., Lin, W., & Jiang, C. (2022). Transformer-based forecasting for auto-scaling microservices in multi-cloud environments. IEEE Access, 10, 88370–88384.
[11] Hasan, M. K., Islam, M. R., & Almogren, A. (2020). Anomaly-aware deep learning-based resource prediction for elastic cloud computing. Sensors, 20(14), 4025.
[12] He, Q., & Jin, H. (2022). Graph neural networks for resource allocation in edge computing. Computer Communications, 180, 140–152.
[13] Hu, Y. C., Patel, M., Sabella, D., Sprecher, N., & Young, V. (2015). Mobile edge computing—A key technology towards 5G. ETSI White Paper, 11(11), 1–16.
[14] Islam, M. R., Hasan, M. K., & Gumaei, A. (2021). Deep LSTM with embedded feature selection for cloud workload prediction. Information Sciences, 562, 136–151.
[15] Kumar, J., Gupta, H., & Buyya, R. (2023). Context-aware auto-scaling for IoT-edge applications. IEEE Internet of Things Journal, 10(5), 4321–4335.
[16] Li, H., Wu, J., Huang, Y., & Hu, X. (2021). Reinforcement learning for dynamic resource provisioning in edge computing. IEEE Transactions on Cloud Computing, 9(3), 1125–1139.
[17] Li, Z., Zhang, Y., & Xu, Y. (2020). Energy-aware container placement in microservice-based edge computing using deep reinforcement learning. IEEE Transactions on Green Communications and Networking, 4(3), 828–842.
[18] Lin, H., & Wang, X. (2021). Multi-objective auto-scaling of containers in edge-cloud systems. Information Systems Frontiers, 23, 1085–1102.
[19] Liu, Z., & Zhang, Q. (2019). A hybrid deep learning framework for workload prediction in cloud systems. Neurocomputing, 361, 177–186.
[20] Mishra, A., Kumar, P., & Verma, S. (2022). Explainable AI for autonomous cloud resource management. ACM Computing Surveys, 55(8), 1–32.
[21] Mohan, N., Ehsan, M., & Wang, Q. (2021). A context-aware policy engine for auto-scaling edge applications. Future Internet, 13(8), 204.
[22] Nguyen, T. T., Nguyen, D. C., & Bhargava, B. (2021). Federated learning for edge-cloud orchestration in IIoT systems. Computer Networks, 194, 108117.
[23] Papageorgiou, E., Loukas, A., & Mylonas, G. (2020). Adaptive auto-scaling using online machine learning. Future Internet, 12(8), 134.
[24] Peng, W., & Wang, Y. (2019). Cost-aware resource management using reinforcement learning in cloud platforms. Future Generation Computer Systems, 95, 70–79.
[25] Qiu, T., Zhao, J., & Yang, S. (2020). Edge computing-enabled intelligent IoT: Performance optimization and adaptive learning. IEEE Network, 34(6), 136–143.
[26] Rahman, M., & Gao, J. (2022). Lightweight deep learning for real-time edge auto-scaling. Computer Communications, 183, 83–92.
[27] Raza, S., & Lee, S. (2021). A comparative study of auto-scaling mechanisms for microservice applications. Sustainable Computing: Informatics and Systems, 31, 100567.
[28] Sharma, A., & Sood, S. K. (2020). Predictive resource provisioning for cloud applications using LSTM networks. Computing, 102(10), 2391–2409.
[29] Singh, R., & Rajan, A. (2023). Online learning-based dynamic scaling for video streaming on edge servers. Multimedia Tools and Applications, 82, 13495–13514.
[30] Sun, Y., & Dong, X. (2020). A survey on adaptive and intelligent auto-scaling strategies. IEEE Transactions on Network and Service Management, 17(4), 2346–2364.
[31] Tang, F., McLaughlin, K., & Sezer, S. (2021). Cloud-native autoscaling with Prometheus and Kubernetes. Software: Practice and Experience, 51(6), 1367–1383.
[32] Thung, G., & Hoon, H. T. (2018). Towards decentralized auto-scaling in fog computing using reinforcement learning. Proceedings of the IEEE Fog World Congress, 1–6.
[33] Torres, J., & Breitbart, Y. (2022). Workload-aware adaptive scaling using online learning. Journal of Cloud Computing, 11, 45.
[34] Wang, L., Li, J., & Xu, H. (2020). Deep learning for workload forecasting in cloud data centers. Journal of Parallel and Distributed Computing, 144, 25–38.
[35] Wang, W., Chen, F., & Zhang, Z. (2021). A survey on federated learning for edge intelligence. ACM Transactions on Internet Technology, 21(2), 1–34.
[36] Wu, J., & Deng, R. H. (2020). AI-powered smart resource provisioning in fog-cloud environments. IEEE Internet Computing, 24(3), 34–42.
[37] Xie, X., & Ding, S. (2021). Contextual deep learning for adaptive scaling in serverless platforms. IEEE Transactions on Cloud Computing, 9(4), 1470–1483.
[38] Yang, J., Li, Z., & Huang, Y. (2019). Real-time scaling in microservices using reinforcement learning. Proceedings of the IEEE International Conference on Cloud Engineering, 140–147.
[39] Zhang, Y., Liu, M., & Tan, Y. (2019). Proactive elasticity in cloud computing: A time-series forecasting approach. Cluster Computing, 22(4), 1459–1474.
[40] Zhang, Z., Xu, X., & Zhou, Y. (2020). AI-driven resource orchestration for service-centric 5G networks. IEEE Wireless Communications, 27(4), 90–97.
