Title: Energy consumption algorithm for unmanned aerial vehicle communication network in marine environment monitoring based on improved deep reinforcement learning

Authors: Jiayi Xin; Hongyan Xing

Addresses: Yantai No. 1 Middle School of Shandong, Erma Road, Zhifu District, Yantai, Shandong, 264025, China ' Shandong Marine Resource and Environment Research Institute, Changjiang Road, Fushan District, Yantai, Shandong, 264025, China

Abstract: Internet of things (IoT) technology and unmanned aerial vehicle (UAV) communication in marine environments can improve task execution efficiency and fault tolerance. Considering the complexity of the marine environment and the energy limitations of UAV network, this paper proposes an energy optimisation algorithm based on an improved deep reinforcement learning (DRL) to minimise the total energy consumption of the network. The incremental gradient descent (ISGD) method is applied to DRL, where sample updates can effectively learn from streaming data in real-time interactions, enhance the interpretability and robustness of the strategy, and supplement the requirements of DRL for adaptive large-scale optimisation of UAV communication networks. The simulation results show that compared with similar algorithms, the proposed ISGD-DRL algorithm can effectively reduce the energy consumption of unmanned aerial vehicle network nodes, has better convergence, and can better achieve the collection and transmission of marine environment monitoring information based on UAV communication network.

Keywords: unmanned aerial vehicle; UAV; energy consumption; deep reinforcement learning; DRL; incremental subgradient descent; ISGD.

DOI: 10.1504/IJAHUC.2026.153837

International Journal of Ad Hoc and Ubiquitous Computing, 2026 Vol.52 No.1, pp.60 - 68

Received: 27 Mar 2025
Accepted: 05 May 2025
Published online: 27 May 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article