Efficient Offloading for Edge Computing in UAV-Assisted Maritime Networks

  • Truong, Thanh Phung; 
  • Do, Tung Son; 
  • Kim, Juyoung; 
  • Choi, Seongjin; 
  • Oh, Junsuk; 
  • ... Cho, Sungrae
Citations

SCOPUS

0

초록

This paper investigates the problem of delay-efficient task execution in UAV-assisted maritime networks. In the pro-posed system, unmanned aerial vehicles (UAVs) are deployed to collect and process computational tasks, either locally or by offloading them to ocean beacon stations (OBSs) equipped with edge computing resources. To minimize the total task execution time while satisfying energy and resource constraints, we formulate a joint optimization problem over offloading ratios, transmit power, and UAV-OBS association. Due to the non-convexity and mixed-variable nature of the problem, a deep reinforcement learning (DRL) framework based on the Deep Deterministic Policy Gradient (DDPG) algorithm is proposed to learn optimal decision policies. Simulation results demonstrate that the proposed approach significantly reduces execution delay compared to baseline methods, highlighting its effectiveness in dynamic and resource-limited maritime environments.

키워드

deep reinforcement learning; efficient offloading; maritime edge networks; unmanned aerial vehicles
제목
Efficient Offloading for Edge Computing in UAV-Assisted Maritime Networks
저자
Truong, Thanh Phung; Do, Tung Son; Kim, Juyoung; Choi, Seongjin; Oh, Junsuk; Cho, Sungrae
DOI
10.1109/ICMIC66299.2025.11257780
발행일
2025
유형
Conference Paper
저널명
ICMIC 2025 - 4th International Conference on Mobile, Military, Maritime IT Convergence: Promoting Ultimate Convergence of Wireless, Military and Maritime Communications in the 6G Era
페이지
337 ~ 340