Title: NOMA-VLC power allocation optimisation assisted by UAV

Authors: Ting Liu; Guangzhao Wang; Jingyu Zhang; Yunshan Sun; Yanqin Li; Teng Fei; Zhanbo Wang

Addresses: Information Engineering College, Tianjin University of Commerce, Tianjin, 300134, China ' Information Engineering College, Tianjin University of Commerce, Tianjin, 300134, China ' China Mobile Communications Group Co., Ltd., Beijing, 100033, China ' Information Engineering College, Tianjin University of Commerce, Tianjin, 300134, China ' Information Engineering College, Tianjin University of Commerce, Tianjin, 300134, China ' Information Engineering College, Tianjin University of Commerce, Tianjin, 300134, China ' Information Engineering College, Tianjin University of Commerce, Tianjin, 300134, China

Abstract: The integration of unmanned aerial vehicles (UAVs), non-orthogonal multiple access (NOMA), and visible light communication (VLC) advances future communication technologies. Despite its potential to overcome spectrum limitations and extend coverage, challenges such as link attenuation and power allocation imbalances hinder performance. This study focuses on UAV-assisted NOMA-VLC systems and proposes dynamic UAV positioning to address limitations of fixed-light-source deployments. A dual marine predator algorithm (DMPA) for power allocation optimisation is introduced, featuring dual-predator reinforcement search, adaptive acceleration factors for improved convergence, and dynamic thresholds based on rate standard deviation and Jain fairness index. Experimental results show that the DMPA outperforms competing schemes in both ideal and obstructed environments. Dynamic UAV positioning enhances signal coverage and system robustness, particularly in obstacle-rich environments.

Keywords: non-orthogonal multiple access; NOMA; visible light communication; VLC; unmanned aerial vehicles; UAVs' swarm intelligence; marine predator algorithm; MPA.

DOI: 10.1504/IJSNET.2026.153828

International Journal of Sensor Networks, 2026 Vol.51 No.1, pp.61 - 74

Received: 11 Oct 2025
Accepted: 20 Oct 2025
Published online: 27 May 2026 *

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