Hierarchical Learning-Based Optimization of Data Flow in Large-Scale Cyber-Physical Systems
Keywords:
Cyber-physical systems, hierarchical learning, data flow optimization, reinforcement learning, distributed systems, adaptive control, edge computingAbstract
Efficient data flow management in large-scale cyber-physical systems (CPS) remains a pressing challenge due to the growing heterogeneity, dynamic operational environments, and the massive volume of interconnected data sources. This paper proposes a hierarchical learning-based optimization framework designed to enhance real-time data flow control across distributed CPS infrastructures. By integrating reinforcement learning within a multi-layered architecture, the proposed approach dynamically adapts to network constraints, prioritizes data packets, and enhances overall throughput while ensuring low latency and fault tolerance. Experimental evaluations demonstrate that hierarchical learning significantly outperforms flat optimization models in terms of scalability, adaptability, and system-wide data efficiency. This work offers critical insights for next-generation intelligent CPS architectures where data-centric decision-making is paramount.
References
1. Zhou, H., Xu, G., Wang, Y.: A Software-Defined Networking Based Dynamic Flow Control in Smart Grids. J Supercomput. 77(4), 3210–3225 (2021)
2. Sun, L., Yang, D.: Edge-Based Optimization in Distributed CPS: An Empirical Approach. J Cloud Comput. 9(2), 115–129 (2020)
3. Lin, X., Wang, Z., Chen, Y.: Deep Reinforcement Learning for Intelligent Routing in Industrial IoT. Neural Comput Appl. 31(8), 2314–2330 (2019)
4. Sharma, K., Park, J.: Real-Time Flow Scheduling in CPS Using Deep Learning Models. Multimed Tools Appl. 79(5), 3381–3400 (2020)
5. Li, B., Zhang, T.: Reinforcement Learning in Fog Computing-Based CPS. Comput J. 63(12), 1825–1838 (2020)
6. Wu, X., Li, F., Ren, K.: Adaptive Routing in Wireless Sensor Networks via HRL. Wireless Pers Commun. 111(3), 2125–2143 (2020)
7. Jin, R., Li, H.: Survey on Learning-Driven CPS Optimization. ACM Trans Embed Comput Syst. 19(4), 35–49 (2020)
8. Qian, Y., Wang, X.: Context-Aware Data Flow in Smart Environments. Future Gener Comput Syst. 102, 527–540 (2020)
9. Kaur, M., Singh, R.: Adaptive Data Flow Models for Industrial CPS. J Ind Inf Integr. 18, 100143 (2020)
10. Rao, N., Bhat, M.: A Layered Learning Strategy for Real-Time CPS. J Syst Archit. 104, 101698 (2020)
11. Tang, L., Xie, L.: Federated Learning in Edge-Centric CPS. J Parallel Distrib Comput. 135, 175–185 (2019)
12. Ahmad, T., Khan, S.: Latency-Aware Routing in Distributed Systems. Clust Comput. 22(6), 13789–13803 (2019)
13. Singh, V., Kumar, P.: Survey of ML in Distributed CPS Networks. Appl Intell. 50(7), 2301–2317 (2020)
14. Zhang, Y., Luo, J.: Cognitive Routing for Data-Centric CPS. J Netw Comput Appl. 137, 38–48 (2020)
15. Wang, L., Zhang, Q.: Multi-Agent Learning in Sensor-Actuator CPS. Inf Sci. 534, 52–68 (2020)
Downloads
Published
Issue
Section
License
Copyright (c) 2022 Imran Amarena, Brandon Ryan (Author)
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
