IEEE ICC 2023 · Conference paper · 2023
Authors: Jiahao You, Ziye Jia, Chao Dong, Lijun He, Yilu Cao, Qihui Wu
Published in: ICC 2023 – IEEE International Conference on Communications, pp. 666–671.
DOI: 10.1109/ICC45041.2023.10279604
IEEE Xplore ↗ | Open preprint on arXiv ↗
Research overview
This paper studies cooperative computation offloading for maritime IoT, using UAVs and vessels to balance task execution time and energy cost.
The problem
Maritime IoT devices generate computation tasks whose demands vary over time. UAVs can support these tasks, but their available energy and computing capacity are limited. This work studies how vessels and UAVs can share the workload.
The approach
We formulate cooperative offloading as a Markov decision process, with execution time and energy cost as the optimization objectives. Lyapunov optimization decomposes the long-term constraints into per-slot problems, which are addressed using Q-learning.
Evaluation
The proposed approach is evaluated through numerical simulations. For the complete model, assumptions, and experimental results, see the published paper or the publicly available preprint.
面向不确定海洋任务的无人机与船舶协同计算卸载
这项工作研究海洋物联网中的协同计算卸载，利用无人机与船舶的计算资源，综合考虑任务执行时间与能耗。
研究问题
海洋物联网设备的计算需求会随时间变化，而无人机的能源和计算能力有限。这项工作研究如何让船舶与无人机共同承担任务，综合考虑执行时间与能耗。
研究方法
论文将协同卸载建模为马尔可夫决策过程，以执行时间和能耗为优化目标。通过 Lyapunov 优化，将长期约束转化为逐时隙问题，再使用 Q-learning 求解。
研究评估
论文通过数值仿真评估所提方法。完整模型、假设和实验结果见发表版本或公开预印本。
中文题名为便于阅读所作的译名，正式引用请使用英文原题。
Citation
[1] J. You, Z. Jia, C. Dong, L. He, Y. Cao, and Q. Wu, “Computation offloading for uncertain marine tasks by cooperation of UAVs and vessels,” in Proc. IEEE Int. Conf. Commun. (ICC), Rome, Italy, 2023, pp. 666–671, doi: 10.1109/ICC45041.2023.10279604.
