GCN and DRL based on dependent task offloading mechanism in edge computing✩

Ruiqi Tong; Shaoyong Guo; Xuesong Qiu; Feng Qi; Dongxiao Yu

doi:10.1016/j.dcan.2025.10.001

›› 2026, Vol. 12 ›› Issue (3) :397 -404. DOI: 10.1016/j.dcan.2025.10.001

research-article

GCN and DRL based on dependent task offloading mechanism in edge computing^✩

Author information +

^a State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications , Beijing 100876, China

^b School of Computer Science and Technology, Shandong University , Qingdao 266510, China

^* E-mail address: syguo@bupt.edu.cn (S. Guo).

Ruiqi Tong received the B.S. degree in Telecommunications Engineering and Management from Beijing University of Posts and Telecommunications, Beijing, China, in 2022. She is currently working toward the Ph.D. degree in Computer Science and Technology at the Beijing University of Posts and Telecommunications, Beijing, China. Her research interests include edge computing, resource scheduling, and Internet of Things (IoT).

Shaoyong Guo is a professor at the School of Computer Science, Beijing University of Posts and Telecommunications. He received the National Science Fund for Excellent Young Scholars in 2023. His research interests include Data Processing Unit (DPU), blockchain application technology, edge intelligence, and so on. He has achieved innovative results, such as power communication network convergence control model and method, network data trusted sandbox privacy sharing service mechanism and technology, and edge security protection technology and mechanism in an open network environment. He is undertaking many key research and development projects and fund projects, and contributed to a number of pioneering standards proposals in ITU-T. The systems and devices developed by him have large-scale application. He was awarded the second prize of Science and Technology Progress Award of Beijing Municipality, Henan Province and Jiangsu Province, and the second prize of Science and Technology Progress Award of Chinese Institute of Electronics, and so on.

Xuesong Qiu received the Ph.D. degree from the Beijing University of Posts and Telecommunications, Beijing, China, in 2000. He is currently a Professor and the Ph.D. Supervisor with the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications. His research interests are mainly in network and service management, edge AI and IoT. He has authored about 100 SCI/EI index papers. He presides over a series of key research projects on network and service management, including the projects supported by the National Natural Science Foundation and the National High Tech Research and Development Program of China.

Feng Qi received the M.E. degree in computer application from Northeastern University, Shenyang, China, in 1996. He is a Professor with Beijing University of Posts and Telecommunications, Beijing, China. He has also written more than ten ITU-T International Standards and Industry Standards, and served as the Vice Chairman of the ITU-T Study Group 4 and Study Group 12. His research interests include communications software, network management, and business intelligence. He has won two National Science and Technology Progress Awards.

Dongxiao Yu received the BS degree from Shandong University in 2006, and the PhD degree from the University of Hong Kong in 2014. He became an associate professor in the School of Computer Science and Technology, Huazhong University of Science and Technology in 2016. He is currently a professor at the School of Computer Science and Technology, Shandong University. His research interests include wireless networks, distributed computing, and graph algorithms.

Show less

History +

PDF (1484KB)

Abstract

Task offloading is critical for optimizing resource allocation in edge computing systems. In practical scenarios, user applications often comprise multiple interdependent tasks, where both task dependencies and parallelism strongly affect offloading decisions. This paper presents a novel dependent task offloading framework for multi-edge server environments. The task offloading problem is formulated as a Markov Decision Process (MDP) to minimize computational delay. Task dependencies are modeled using a Directed Acyclic Graph (DAG), and a Graph Convolutional Network (GCN) encoder is employed to extract DAG features as inputs for a Deep Reinforcement Learning (DRL) model. The proposed DRL-based method applies the Proximal Policy Optimization (PPO) algorithm to simultaneously select subtasks and determine their offloading decisions. Experimental evaluations across varying numbers of subtasks confirm the effectiveness of the approach, demonstrating superior performance compared to state-of-the-art solutions.

Keywords

Task offloading / Edge computing / Graph neural network / Deep reinforcement learning

Cite this article

Download citation ▾

Ruiqi Tong, Shaoyong Guo, Xuesong Qiu, Feng Qi, Dongxiao Yu. GCN and DRL based on dependent task offloading mechanism in edge computing^✩. , 2026, 12 (3) : 397-404 DOI:10.1016/j.dcan.2025.10.001

登录浏览全文

4963

注册一个新账户忘记密码

CRediT authorship contribution statement

Ruiqi Tong: Writing -- review & editing, Writing -- original draft, Validation, Methodology, Investigation, Formal analysis. Shaoyong Guo: Writing -- review & editing, Supervision, Funding acquisition. Xuesong Qiu: Supervision, Funding acquisition. Feng Qi: Supervision. Dongxiao Yu: Supervision.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Shaoyong Guo reports financial support was provided by National Natural Science Foundation of China (62322103). Shaoyong Guo reports financial support was provided by the Fund of Central University Basic Research Project (2023ZCTH11). Xuesong Qiu reports financial support was provided by Beijing Natural Science Foundation (4232009). Reports a relationship with that includes: Has patent pending to. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

This work is supported by the National Natural Science Foundation of China (62322103), Beijing Natural Science Foundation (4232009), and the Fund of Central University Basic Research Projects (2023ZCTH11).

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]	Z. Cai, X. Zheng, A private and efficient mechanism for data uploading in smart cyber—physical systems, IEEE Trans. Netw. Sci. Eng. 7 (2) (2020) 766-775.

[2]	Z. Cai, Z. Duan, W. Li, Exploiting multi—dimensional task diversity in distributed auctions for mobile crowdsensing, IEEE Trans. Mob. Comput. 20 (8) (2021) 2576-2591.

[3]	S. Wang, X. Song, H. Xu, T. Song, G. Zhang, Y. Yang, Joint offloading decision and resource allocation in vehicular edge computing networks, Digit. Commun. Netw. 11 (1) (2025) 71-82.

[4]	Y. Gong, D. Yu, H. Yao, X. Cheng, A. Nallanathan, G.K. Karagiannidis, Multi—modal learning based multi—task offloading schemes for satellite—ground integrated networks, IEEE Trans. Wirel. Commun. 24 (7) (2025) 5635-5647.

[5]	S. Xia, Z. Yao, G. Wu, Y. Li, Distributed offloading for cooperative intelligent transportation under heterogeneous networks, IEEE Trans. Intell. Transp. Syst. 23 (9) (2022) 16701-16714.

[6]	F. Song, Y. Ma, I. You, H. Zhang, Smart collaborative evolvement for virtual group creation in customized industrial iot, IEEE Trans. Netw. Sci. Eng. 10 (5) (2023) 2514-2524.

[7]	F. Song, Y. Ma, Z. Yuan, I. You, G. Pau, H. Zhang, Exploring reliable decentralized networks with smart collaborative theory, IEEE Commun. Mag. 61 (8) (2023) 44-50.

[8]	P. Porambage, J. Okwuibe, M. Liyanage, M. Ylianttila, T. Taleb, Survey on multi—access edge computing for internet of things realization, IEEE Commun. Surv. Tutor. 20 (4) (2018) 2961-2991.

[9]	Z. Cai, T. Shi, Distributed query processing in the edge—assisted iot data monitoring system, IEEE Internet Things J. 8 (16) (2021) 12679-12693.

[10]	T. Zhu, T. Shi, J. Li, Z. Cai, X. Zhou, Task scheduling in deadline—aware mobile edge computing systems, IEEE Internet Things J. 6 (3) (2019) 4854-4866.

[11]	Z. Tang, J. Lou, F. Zhang, W. Jia, Dependent task offloading for multiple jobs in edge computing, in: 2020 29th International Conference on Computer Communications and Networks (ICCCN), IEEE, 2020, pp. 1-9.

[12]	Q. Luo, S. Hu, C. Li, G. Li, W. Shi, Resource scheduling in edge computing: a survey, IEEE Commun. Surv. Tutor. 23 (4) (2021) 2131-2165.

[13]	Y. Huang, Multimedia tasks—oriented edge computing offloading scheme based on graph neural network in vehicular networks, IEEE Access 13 (2025) 9780-9791.

[14]	H. Wu, Y. Sun, K. Wolter, Energy—efficient decision making for mobile cloud offloading, IEEE Trans. Cloud Comput. 8 (2) (2020) 570-584.

[15]	J. Xu, L. Chen, P. Zhou, Joint service caching and task offloading for mobile edge computing in dense networks, in: IEEE INFOCOM 2018 — IEEE Conference on Computer Communications, IEEE, 2018, pp. 207-215.

[16]	Z. Han, H. Tan, X.—Y. Li, S.H.—C. Jiang, Y. Li, F.C.M. Lau, Ondisc: online latency—sensitive job dispatching and scheduling in heterogeneous edge—clouds, IEEE/ACM Trans. Netw. 27 (6) (2019) 2472-2485.

[17]	L. Xiao, X. Lu, T. Xu, X. Wan, W. Ji, Y. Zhang, Reinforcement learning—based mobile offloading for edge computing against jamming and interference, IEEE Trans. Commun. 68 (10) (2020) 6114-6126.

[18]	J. Wang, J. Hu, G. Min, W. Zhan, A.Y. Zomaya, N. Georgalas, Dependent task offloading for edge computing based on deep reinforcement learning, IEEE Trans. Comput. 71 (10) (2022) 2449-2461.

[19]	S. Shen, Y. Han, X. Wang, Y. Wang, Computation offloading with multiple agents in edge—computing—supported iot, ACM Trans. Sens. Netw. 16 (1) (2019) 1-27.

[20]	B. Gu, M. Alazab, Z. Lin, X. Zhang, J. Huang, Ai—enabled task offloading for improving quality of computational experience in ultra dense networks, ACM Trans. Internet Technol. 22 (3) (2022) 1-17.

[21]	C. Shu, Z. Zhao, Y. Han, G. Min, H. Duan, Multi—user offloading for edge computing networks: a dependency—aware and latency—optimal approach, IEEE Internet Things J. 7 (3) (2020) 1678-1689.

[22]	J.D. Ullman, Np—complete scheduling problems, J. Comput. Syst. Sci. 10 (3) (1975) 384-393.

[23]	H. Liao, X. Li, D. Guo, W. Kang, J. Li, Dependency—aware application assigning and scheduling in edge computing, IEEE Internet Things J. 9 (6) (2022) 4451-4463.

[24]	C. Feng, P. Han, X. Zhang, Q. Zhang, Y. Liu, L. Guo, Dependency—aware task reconfiguration and offloading in multi—access edge cloud networks, IEEE Trans. Mob. Comput. 23 (10) (2024) 9271-9288.

[25]	C. Shu, Z. Zhao, Y. Han, G. Min, Dependency—aware and latency—optimal computation offloading for multi—user edge computing networks, in: 2019 16th Annual IEEE International Conference on Sensing, Communication, and Networking (SECON), IEEE, 2019, pp. 1-9.

[26]	S. Liu, Y. Yu, X. Lian, Y. Feng, C. She, P.L. Yeoh, L. Guo, B. Vucetic, Y. Li, Dependent task scheduling and offloading for minimizing deadline violation ratio in mobile edge computing networks, IEEE J. Sel. Areas Commun. 41 (2) (2023) 538-554.

[27]	N. Sharma, A. Ghosh, R. Misra, S.K. Das, Deep meta q—learning based multi—task offloading in edge—cloud systems, IEEE Trans. Mob. Comput. 23 (4) (2024) 2583-2598.

[28]	J. Yuan, H. Xiao, Z. Shen, T. Zhang, J. Jin, Elect: energy—efficient intelligent edge—cloud collaboration for remote iot services, Future Gener. Comput. Syst. 147 (2023) 179-194.

[29]	X. Zhou, S. Ge, P. Liu, T. Qiu, Dag—based dependent tasks offloading in mec—enabled iot with soft cooperation, IEEE Trans. Mob. Comput. 23 (6) (2024) 6908-6920.

[30]	J. Chen, Y. Yang, C. Wang, H. Zhang, C. Qiu, X. Wang, Multitask offloading strategy optimization based on directed acyclic graphs for edge computing, IEEE Internet Things J. 9 (12) (2022) 9367-9378.

[31]	Z. Cao, X. Deng, S. Yue, P. Jiang, J. Ren, J. Gui, Dependent task offloading in edge computing using gnn and deep reinforcement learning, IEEE Internet Things J. 11 (12) (2024) 21632-21646.

[32]	T.N. Kipf, M. Welling, Semi—supervised classification with graph convolutional networks, arXiv preprint, arXiv:1609.02907, https://doi.org/10.48550/arXiv.1609.02907.

[33]	J. Schulman, F. Wolski, P. Dhariwal, A. Radford, O. Klimov, Proximal policy optimization algorithms, arXiv preprint, arXiv:1707.06347, https://doi.org/10.48550/arXiv.1707.06347.

[34]	S. Zhao, X. Dai, I. Bate, A. Burns, W. Chang, Dag scheduling and analysis on multiprocessor systems: exploitation of parallelism and dependency, in: 2020 IEEE Real—Time Systems Symposium (RTSS), IEEE, 2020, pp. 128-140.

[35]	S. Zhao, X. Dai, I. Bate, Dag scheduling and analysis on multi—core systems by modelling parallelism and dependency, IEEE Trans. Parallel Distrib. Syst. 33 (12) (2022) 4019-4038.