Reinforcement Learning Evaluation for Dependent Task Offloading in Mobile Edge Computing Systems
DOI:
10.33395/sinkron.v10i4.16636Keywords:
Deep reinforcement learning, Directed acyclic graph, Mobile edge computing, Quality of experience, Task offloadingAbstract
Dependent-task directed acyclic graphs require each task to execute locally or offload to a mobile edge computing server, coupling latency, communication cost, and device energy. This study reports a controlled evaluation of deep reinforcement learning for offloading under a common simulator, dataset, and train/validation/final-test protocol. Baselines comprise a recurrent policy with proximal policy optimization (PPO-LSTM) and a double deep Q-network; extensions comprise a Transformer policy, discrete soft actor-critic, and stabilized PPO with advantage normalization, scheduled optimization, divergence-based early stopping, and orthogonal initialization. Agents and five heuristics are evaluated on final-test graphs of ten to fifty tasks under latency-oriented and energy-aware quality of experience (QoE) objectives, with five seeds and Holm-corrected tests. Stabilized PPO attained mean QoE of 0.31–0.43 across task-size and bandwidth analyses, though not at every size or bandwidth mean, exceeding Heterogeneous Earliest Finish Time (HEFT) by 0.05 to 0.15 absolute, yet its gain over PPO-LSTM remained about 0.001 and was not Holm-significant at any bandwidth. Both PPO variants reached the validation threshold at median update five, against ten for the Transformer. Ablation identified no single driver: only removing advantage normalization produced a Holm-significant drop, and only under the energy-aware objective. The Transformer remained competitive, whereas discrete soft actor-critic reduced energy rather than latency. Matching the recurrent baseline in size and inference time, stabilization remains viable on resource-constrained devices but cannot replace retraining on much slower links.
Downloads
References
Ahmed, M., Raza, S., Ahmad, H., Khan, W. U., Xu, F., & Rabie, K. (2024). Deep reinforcement learning approach for multi-hop task offloading in vehicular edge computing. Engineering Science and Technology, an International Journal, 59. https://doi.org/10.1016/j.jestch.2024.101854
Christodoulou, P. (2019). Soft Actor-Critic for Discrete Action Settings. http://arxiv.org/abs/1910.07207
Cui, Y., Li, H., Zhang, D., Zhu, A., Li, Y., & Qiang, H. (2023). Multiagent Reinforcement Learning-Based Cooperative Multitype Task Offloading Strategy for Internet of Vehicles in B5G/6G Network. IEEE Internet of Things Journal, 10(14), 12248–12260. https://doi.org/10.1109/JIOT.2023.3245721
Gao, Z., Yang, L., & Dai, Y. (2023). Fast Adaptive Task Offloading and Resource Allocation via Multiagent Reinforcement Learning in Heterogeneous Vehicular Fog Computing. IEEE Internet of Things Journal, 10(8), 6818–6835. https://doi.org/10.1109/JIOT.2022.3228246
Hochreiter, S., & Schmidhuber, J. (1997). Long Short-Term Memory. Neural Computation, 9(8), 1735–1780. https://doi.org/10.1162/neco.1997.9.8.1735
Hu, X., & Huang, Y. (2022). Deep reinforcement learning based offloading decision algorithm for vehicular edge computing. PeerJ Computer Science, 8. https://doi.org/10.7717/PEERJ-CS.1126
Huang, J., Xiao, H., Xiao, L., & Gu, Z. (2026). Deep Reinforcement Learning-Based DAG Task Scheduling for Heterogeneous Edge–Cloud Systems. Journal of Circuits, Systems and Computers. https://doi.org/10.1142/S0218126626502361
Ju, X., Su, S., Xu, C., & Wang, H. (2023). Computation offloading and tasks scheduling for the internet of vehicles in edge computing: A deep reinforcement learning-based pointer network approach. Computer Networks, 223, 109572. https://doi.org/10.1016/j.comnet.2023.109572
Ke, H., Wang, H., & Sun, H. (2022). Multi-Agent Deep Reinforcement Learning-Based Partial Task Offloading and Resource Allocation in Edge Computing Environment. Electronics (Switzerland), 11(15). https://doi.org/10.3390/electronics11152394
Li, Y.-S., & Gau, R.-H. (2024). Transformer-Assisted Deep Reinforcement Learning for Distributed Latency-Sensitive Task Offloading in Mobile Edge Computing. ICC 2024 - IEEE International Conference on Communications, 2944–2949. https://doi.org/10.1109/ICC51166.2024.10622508
Li, Z., Fu, Y., Tian, M., & Li, C. (2024). Cooperative sensing, communication and computation resource allocation in mobile edge computing-enabled vehicular networks. Journal of Information and Intelligence, 2(4), 339–354. https://doi.org/10.1016/j.jiixd.2024.02.006
Liu, C., Wang, H., Zhao, M., Liu, J., Zhao, X., & Yuan, P. (2024). Dependency-aware online task offloading based on deep reinforcement learning for IoV. Journal of Cloud Computing, 13(1). https://doi.org/10.1186/s13677-024-00701-0
Ma, B., Xu, Y., Pan, Y., Liu, S., & Li, C. (2024). A multi-user mobile edge computing task offloading and trajectory management based on proximal policy optimization. Peer-to-Peer Networking and Applications, 17(6), 4210–4229. https://doi.org/10.1007/s12083-024-01796-7
Moon, S., & Lim, Y. (2022). Federated Deep Reinforcement Learning Based Task Offloading with Power Control in Vehicular Edge Computing. Sensors, 22(24). https://doi.org/10.3390/s22249595
Nie, X., Yan, Y., Zhou, T., Chen, X., & Zhang, D. (2023). A Delay-Optimal Task Scheduling Strategy for Vehicle Edge Computing Based on the Multi-Agent Deep Reinforcement Learning Approach. Electronics (Switzerland), 12(7). https://doi.org/10.3390/electronics12071655
Pang, S., Hou, L., Gui, H., He, X., Wang, T., & Zhao, Y. (2024). Multi-mobile vehicles task offloading for vehicle-edge-cloud collaboration: A dependency-aware and deep reinforcement learning approach. Computer Communications, 213, 359–371. https://doi.org/10.1016/j.comcom.2023.11.013
Schulman, J., Wolski, F., Dhariwal, P., Radford, A., & Klimov, O. (2017). Proximal Policy Optimization Algorithms. http://arxiv.org/abs/1707.06347
Shi, W., Chen, L., & Zhu, X. (2023). Task Offloading Decision-Making Algorithm for Vehicular Edge Computing: A Deep-Reinforcement-Learning-Based Approach. Sensors, 23(17). https://doi.org/10.3390/s23177595
Sivasakthi, D. A., & Gunasekaran, R. (2022). QoE‐aware mobile computation offloading in mobile edge computing. Concurrency and Computation: Practice and Experience, 34(11). https://doi.org/10.1002/cpe.6853
Sun, X., Hu, Y., Gao, X., & Wang, H. (2025). A cooperative multi-agent optimization approach for task offloading in vehicular edge computing systems. Discover Computing, 28(1). https://doi.org/10.1007/s10791-025-09887-6
Tang, T., Li, C., & Liu, F. (2023). Collaborative cloud-edge-end task offloading with task dependency based on deep reinforcement learning. Computer Communications, 209, 78–90. https://doi.org/10.1016/j.comcom.2023.06.021
Topcuoglu, H., Hariri, S., & Wu, M.-Y. (2002). Performance-Effective and Low-Complexity Task Scheduling for Heterogeneous Computing.
Uddin, A., Sakr, A. H., & Zhang, N. (2024). Prioritized Task Offloading in Vehicular Edge Computing Using Deep Reinforcement Learning. 2024 IEEE 99th Vehicular Technology Conference (VTC2024-Spring), 1–6. https://doi.org/10.1109/VTC2024-Spring62846.2024.10683050
Van Hasselt, H., Guez, A., & Silver, D. (2016). Deep Reinforcement Learning with Double Q-Learning. Retrieved www.aaai.org
Wang, J., Hu, J., Min, G., Zhan, W., Zomaya, A. Y., & Georgalas, N. (2021). Dependent Task Offloading for Edge Computing based on Deep Reinforcement Learning. Retrieved https://github.com/linkpark/RLTaskOffloading
Wu, Z., Jia, Z., Pang, X., & Zhao, S. (2024). Deep Reinforcement Learning-Based Task Offloading and Load Balancing for Vehicular Edge Computing. Electronics (Switzerland), 13(8). https://doi.org/10.3390/electronics13081511
Yang, W., Liu, Z., Liu, X., & Ma, Y. (2024). Deep reinforcement learning-based low-latency task offloading for mobile-edge computing networks. Applied Soft Computing, 166, 112164. https://doi.org/10.1016/j.asoc.2024.112164
Zhang, Y., Chen, J., Zhou, Y., Yang, L., He, B., & Yang, Y. (2022). Dependent task offloading with energy-latency tradeoff in mobile edge computing. IET Communications, 16(17), 1993–2001. https://doi.org/10.1049/cmu2.12454
Zhao, H., Li, Y., Pang, Z., & Ma, Z. (2025). Federated Multi-Agent DRL for Task Offloading in Vehicular Edge Computing. Electronics (Switzerland), 14(17). https://doi.org/10.3390/electronics14173501
Zhou, J., Liang, J., Zhao, L., Wan, S., Cai, H., & Xiao, F. (2025). Latency-Energy Efficient Task Offloading in the Satellite Network-Assisted Edge Computing via Deep Reinforcement Learning. IEEE Transactions on Mobile Computing, 24(4), 2644–2659. https://doi.org/10.1109/TMC.2024.3502643
Downloads
|
How to Cite
Issue
Section
License
Copyright (c) 2026 Lubna Thair, Awos Kh. Ali
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Muhammad Khoiruddin harahap
Google Scholar 
Moraref
PKP Index
Garuda Kemdikbud
Indonesia OneSearch
OCLC Worldcat
Dimensions
Index Copernicus
Scilit
