On the Performance of Unmanned Aerial Vehicles with MIMO VLC

Hosein Zarini, Amir Mohammadi, Maryam Farajzadeh Dehkordi, Mohammad Robat Mili, Bardia Safaei, Ali Movaghar, Mérouane Debbah, Jorswieck, EduardarXiv 2024 · DOI · arXiv

Abstract

This paper centers around a multiple-input-multiple-output (MIMO) visible light communication (VLC) system, where an unmanned aerial vehicle (UAV) benefits from a light emitting diode (LED) array to serve photo-diode (PD)-equipped users for illumination and communication simultaneously. Concerning the battery limitation of the UAV and considerable energy consumption of the LED array, a hybrid dimming control scheme is devised at the UAV that effectively controls the number of glared LEDs and thereby mitigates the overall energy consumption. To assess the performance of this system, a radio resource allocation problem is accordingly formulated for jointly optimizing the motion trajectory, transmit beamforming and LED selection at the UAV, assuming that channel state information (CSI) is partially available. By reformulating the optimization problem in Markov decision process (MDP) form, we propose a soft actor-critic (SAC) mechanism that captures the dynamics of the problem and optimizes its parameters. Additionally, regarding the high mobility of the UAV and thus remarkable rearrangement of the system, we enhance the trained SAC model by integrating a meta-learning strategy that enables more adaptation to system variations. By defining energy efficiency as a trade-off between the data rate and power consumption, simulations verify that upgrading a single-LED UAV by an array of 10 LEDs, exhibits 47% and 34% improvements in data rate and energy efficiency, albeit at the expense of 8% more power consumption.

Introduction

Visible Light Communication (VLC) is emerging as a high-bandwidth alternative to traditional RF systems, utilizing LEDs for simultaneous illumination and data transmission. While Unmanned Aerial Vehicles (UAVs) have been proposed as mobile VLC base stations, most research has focused on single-LED configurations. This paper investigates the transition to Multiple-Input Multiple-Output (MIMO) VLC using LED arrays on UAVs, addressing the critical challenges of power constraints and high system dynamism.

System Model and Hybrid Dimming Control

The proposed system features a UAV equipped with an LED array serving multiple photo-diode (PD)-equipped users. To manage the significant power consumption of LED arrays alongside the UAV's battery limitations, a hybrid dimming control scheme is introduced. This scheme jointly manages DC-bias (analog dimming) and the number of active LEDs (spatial dimming), ensuring illumination requirements are met without wasting energy on unnecessary "glaring" LEDs.

System model of the UAV-MIMO-VLC architecture including the hybrid dimming control and LED selection process.

The resource allocation problem is formulated to jointly optimize the UAV's flight trajectory, transmit beamforming, and LED selection under conditions of partial Channel State Information (CSI).

Meta-Learning Enhanced Soft Actor-Critic

Given the NP-hard nature of the optimization problem and the high mobility of the UAV, the authors employ a Deep Reinforcement Learning (DRL) approach. A Soft Actor-Critic (SAC) agent is utilized for its stability and exploration capabilities. To enhance adaptability to rapid environmental changes, such as varying user placements or trajectories, Model-Agnostic Meta-Learning (MAML) is integrated into the SAC framework (Meta-SAC). This allows the model to "learn how to learn," enabling fast adaptation to new tasks with minimal fine-tuning.

Performance Evaluation and Key Results

Simulations were conducted with 30 users and UAVs equipped with 10 to 20 LEDs. The results demonstrate that moving from a single-LED to a 10-LED MIMO array yields a 47% improvement in data rate and a 34% boost in energy efficiency, with only an 8% increase in total power consumption.

Convergence performance of Meta-SAC compared to standard SAC and greedy baselines.

As shown in the convergence plots, Meta-SAC reaches a stable, high-reward state significantly faster and with less variance than traditional SAC or greedy methods. Furthermore, Meta-SAC achieved an energy efficiency score of 3.41 in the 10-LED scenario, significantly outperforming standard SAC (2.62) and PPO (2.11). Notably, the Meta-SAC approach consumed roughly 65% less power on average than standard SAC due to its ability to find efficient flight and transmission paths more rapidly.

Conclusion and Future Work

This study confirms that MIMO VLC arrays significantly enhance UAV communication performance and energy efficiency. The Meta-SAC algorithm provides the necessary robustness and adaptability for dynamic aerial missions. Future research will address Non-Line-of-Sight (NLoS) signal components in complex environments and scenarios involving highly mobile users, as well as real-world hardware constraints such as heat dissipation and driver weight.

Summary updated

Abstract

This paper centers around a multiple-input-multiple-output (MIMO) visible light communication (VLC) system, where an unmanned aerial vehicle (UAV) benefits from a light emitting diode (LED) array to serve photo-diode (PD)-equipped users for illumination and communication simultaneously. Concerning the battery limitation of the UAV and considerable energy consumption of the LED array, a hybrid dimming control scheme is devised at the UAV that effectively controls the number of glared LEDs and thereby mitigates the overall energy consumption. To assess the performance of this system, a radio resource allocation problem is accordingly formulated for jointly optimizing the motion trajectory, transmit beamforming and LED selection at the UAV, assuming that channel state information (CSI) is partially available. By reformulating the optimization problem in Markov decision process (MDP) form, we propose a soft actor-critic (SAC) mechanism that captures the dynamics of the problem and optimizes its parameters. Additionally, regarding the high mobility of the UAV and thus remarkable rearrangement of the system, we enhance the trained SAC model by integrating a meta-learning strategy that enables more adaptation to system variations. By defining energy efficiency as a trade-off between the data rate and power consumption, simulations verify that upgrading a single-LED UAV by an array of 10 LEDs, exhibits 47% and 34% improvements in data rate and energy efficiency, albeit at the expense of 8% more power consumption.
BibTeX
@misc{zarini2024performance,
  author = {Hosein Zarini and Amir Mohammadi and Maryam Farajzadeh Dehkordi and Mohammad Robat Mili and Bardia Safaei and Ali Movaghar and Mérouane Debbah and Jorswieck, Eduard},
  title = {{On the Performance of Unmanned Aerial Vehicles with MIMO VLC}},
  year = {2024},
  eprint = {2405.11161},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2405.11161},
}
Figure