Delay- and Energy-Efficient Task Offloading in Active RIS-Assisted UAV-VEC Network: A Heterogeneous Agent Proximal Policy Optimization Approach
IEEE Transactions on Mobile Computing · 2026-12 · 25 (12) · 22900-22916
DOI: 10.1109/tmc.2026.3725111 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Mobile Computing, 2026-12. Verfasst von Shujuan Wang und 2 weiteren Autor:innen. Die Arbeit nennt 43 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Yunnan Provincial Department of Science and Technology, Fundamental Research Funds for the Central Universities.
Abstract (Originalsprache)
UAV-assisted task offloading is a promising technique for delay sensitive and computation intensive applications in vehicular edge computing (VEC). Nevertheless, practical deployment faces several challenges. The randomly generated computation tasks cause bursty, imbalanced network workload and degrade RSU responsiveness. Moreover, complex VEC communication environments hinder efficient transmission among vehicles, UAVs and RSUs. Besides, UAV energy constraints impair service performance and prevent optimal offloading decisions. This paper proposes a joint optimization algorithm based on Heterogeneous Agent Proximal Policy Optimization (HAPPO). A three tier multi UAV collaborative offloading architecture is developed to support task migration among RSUs for real time load balancing and balanced communication energy resource consumption. A specifically devised communication strategy, which combines the active reconfigurable intelligent surface (RIS) and the multiple input multiple output (MIMO) technologies, is proposed to combat the lossy, unstable channels in VEC. The phase and power control of each RIS is intelligently guided so that the wireless channel can be jointly optimized. Furthermore, a solar-powering energy model is devised to provide sustainable service capability for UAVs. The coupling and conflicts between different trajectory plans aiming at varied objectives are elegantly resolved. Unlike prior works that treat RIS and UAV as independent components, the proposed architecture achieves synergistic co-design in which active RIS mounted on UAVs simultaneously compensates multiplicative path loss in cascaded multi-hop channels and exploits UAV mobility to steer beams toward dynamically moving vehicles, a coupling that passive RIS cannot replicate. Simulation results show that the proposed algorithm achieves satisfying performance in terms of convergence, delay and energy consumption under varied circumstances.
Förderung
- National Natural Science Foundation of China
- Yunnan Provincial Department of Science and Technology
- Fundamental Research Funds for the Central Universities
- ZTE Industry-University-Institute Cooperation Funds
- Zhongguancun XinXi Disruptive Technology Innovation Foundation
Zitieren
APA
Wang, S., Jin, G., & Wang, F. (2026). Delay- and Energy-Efficient Task Offloading in Active RIS-Assisted UAV-VEC Network: A Heterogeneous Agent Proximal Policy Optimization Approach. IEEE Transactions on Mobile Computing, 25(12), 22900-22916. https://doi.org/10.1109/tmc.2026.3725111
Vancouver
Wang S, Jin G, Wang F. Delay- and Energy-Efficient Task Offloading in Active RIS-Assisted UAV-VEC Network: A Heterogeneous Agent Proximal Policy Optimization Approach. IEEE Transactions on Mobile Computing. 2026;25(12):22900-22916. doi:10.1109/tmc.2026.3725111
BibTeX
@article{Wang2026,
author = {Wang, Shujuan and Jin, Guangxiong and Wang, Fanggang},
title = {Delay- and Energy-Efficient Task Offloading in Active RIS-Assisted UAV-VEC Network: A Heterogeneous Agent Proximal Policy Optimization Approach},
journal = {IEEE Transactions on Mobile Computing},
year = {2026},
volume = {25},
number = {12},
pages = {22900--22916},
doi = {10.1109/tmc.2026.3725111}
}
EUVIMED