A Trusted Edge Scheduling Method for Computing Power Network Based on Deep Reinforcement Learning
DOI:
https://doi.org/10.4108/eetsis.13053Keywords:
deep reinforcement learning, graph neural network, blockchain, smart contract, reverse proxyAbstract
In multi-party collaborative scenarios where computing power networks and edge intelligence work in synergy, trusted edge scheduling methods optimize resource allocation through intelligent scheduling, ensuring that computing tasks of each participant are executed on independent edge nodes, thus effectively improving computing efficiency and resource utilization. However, distributed edge nodes and their supporting communication networks usually require independent deployment and operation and maintenance, making efficient scheduling a critical challenge. This paper proposes a trusted edge scheduling method based on deep reinforcement learning, which realizes privacy protection through smart contract technology. Computing and network resources are stored on the blockchain, scheduling is performed using the Edge Routing Model (ERM) in smart contracts, and the hash values of scheduling results are recorded on the blockchain. ERM aims to meet users’ computing demands while selecting the optimal routing strategy to maximize network bandwidth utilization. The method adopts deep reinforcement learning with a prioritized sampling mechanism to constrain the action space, and combines graph neural networks to learn features of network topology, thereby enhancing the generalization ability. Experimental results show that, compared with traditional scheduling models, ERM can schedule edge resources more efficiently under both homogeneous and heterogeneous topologies and exhibits favorable scalability.
References
[1] Saleh A, Joshi P, Rathore R S, et al. Trust-aware routing mechanism through an edge node for IoT-enabled sensor networks. Sensors. 2022; 22(20):7820.
[2] Li J, Zhang H, Li S, et al. BD-TTS: A blockchain and DRL-based framework for trusted task scheduling in edge computing. Computer Networks. 2024; 251:110609.
[3] Kang K, Ding D, Xie H, et al. Adaptive DRL-based task scheduling for energy-efficient cloud computing. IEEE Transactions on Network and Service Management. 2022; 19(4):4948-4961.
[4] Sun P, Guo Z, Li J, et al. Enabling Scalable routing in software defined networks With deep reinforcement learning on critical nodes. IEEE/ACM Transactions on Networking. 2022; 30(2): 629-640.
[5] Avan A, Azim A, Mahmoud Q H. A state-of-the-art review of task scheduling for edge computing: A delay-sensitive application perspective. Electronics. 2023; 12(12): 2599.
[6] Tang X, Cao C, Wang Y, et al. Computing power network: The architecture of convergence of computing and networking towards 6G requirement. China Communications. 2021; 18(2):175-185.
[7] Moerland T M, Broekens J, Plaat A, et al. Model-based reinforcement learning: A survey. Foundations and Trends in Machine Learning. 2023; 16(1):1-118.
[8] Lyu L, Shen Y, Zhang S, The advance of reinforcement learning and deep reinforcement learning. In: 2022 IEEE International Conference on Electrical Engineering Big Data and Algorithms (EEBDA); Changchun, China. 2022. p.644-648.
[9] Scarselli F, Gori M, Tsoi A C, et al. The graph neural network model. IEEE Transactions on Neural Networks. 2008; 20(1):61-80.
[10] Zhou J, Cui G, Hu S, et al. Graph neural networks: A review of methods and applications. AI Open. 2020; 1:57-81.
[11] Masruroh S U, Fiade A, Iman M F, et al. Performance evaluation of routing protocol RIPv2, OSPF, EIGRP with BGP. In: 2017 International Conference on Innovative and Creative Information Technology (ICITech); Salatiga, Indonesia. 2017. p. 1-7.
[12] Rajeswari A R, Network A. Routing protocols: a comparative study. Recent trends in communication networks. 2020:225.
[13] Apache Software Foundation. Forward and Reverse Proxies. https://httpd.apache.org/docs/2.0/mod/mod_proxy.html, last accessed August 1, 2024.
[14] Kim M, Park J. Learning collaborative policies to solve np-hard routing problems. Advances in Neural Information Processing Systems. 2021; 34:10418-10430.
[15] Huang K, Zhai J, Zheng Z, et al. Understanding and bridging the gaps in current GNN performance optimizations. In: Proceedings of the 26th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming. 2021. p. 119-132.
[16] Sherstinsky A. Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network. Physica D: Nonlinear Phenomena. 2020; 404:132306.
[17] Fan J, Wang Z, Xie Y, et al. A theoretical analysis of deep Q-learning. In: Learning for dynamics and control. PMLR. 2020. p. 486-489.
[18] Fraunholz D, Reti D, Duque Anton S, et al. Cloxy: A context-aware deception-as-a-service reverse proxy for web services. In: Proceedings of the 5th ACM workshop on moving target defense. 2018. p. 40-47.
[19] Knight S, Nguyen H, Falkner N, Bowden R, et al. The internet topology zoo. IEEE Journal on Selected Areas in Communications. 2011; 29(9):1765–1775.
[20] Garg S, Gupta S, Srivastava V. Resource allocation in iot networks via DQN and PPO algorithm. In: 2025 2nd International Conference on Computational Intelligence, Communication Technology and Networking (CICTN). 2025. p. 23-29.
[21] Anbazhagan K, Kumar R, Thilagavathy R, et al. Shortest job first with Gateway-based resource management strategy for fog enabled cloud computing. In: 2024 4th International Conference on Data Engineering and Communication Systems (ICDECS). 2024. p. 1-6.
[22] Alizadeh M, Yang S, Sharif M, et al. pFabric: Minimal near-optimal datacenter transport. In: Proceedings of ACM SIGCOMM. 2013. p. 435–446.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Yong Yan, Ronghao Liu, Xu Liu, Ao Xiong, Zuhao Wang
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
This is an open access article distributed under the terms of the CC BY-NC-SA 4.0, which permits copying, redistributing, remixing, transformation, and building upon the material in any medium so long as the original work is properly cited.
