Smart 6G Sky for Green Mobile IOT Networks

TL;DR

This paper proposes an AI-driven approach using deep reinforcement learning to optimize UAV-assisted IoT networks for 6G, balancing data rate, energy harvesting, and energy consumption for greener, more flexible connectivity.

Contribution

It introduces a novel application of deep deterministic policy gradient (DDPG) for joint optimization in UAV-assisted IoT networks, outperforming traditional methods.

Findings

DDPG effectively balances data rate, energy harvesting, and energy consumption.

The proposed method outperforms traditional mathematical schemes.

The approach is applicable to mobile IoT devices in UAV-assisted networks.

Abstract

6G is envisioned to connect everything and yet to be a hundred times more energy efficient than the 5G. Thanks for its ability to use sources of ambient energy, energy harvesting (EH) is promising in alleviating the challenge of meeting such conflicting demands. Moreover, when it comes to the Internet of things (IoT), one of the foundations for enabling connecting everything, the need for EH may become inevitable. IoT involves connecting not only devices that are large in number, but also hard to reach. The good news, nevertheless, is that the unmanned aerial vehicle (UAV), owning to its flexibility and ease of deployment is emerging to offer communication services when infrastructure is lacking. Merging the UAV and IoT is of quite interest as the former could not just enable flexible connectivity for the IoT but also powering them in spite of any restrictions. However, managing the UAV assisted IoT resources to meet certain data communications and EH quality measures while keeping the UAV consumed energy minimized is a major challenge as this corresponds to a non-convex optimization problem. Things, obviously, become even worse when the IoT network devices are mobile. Owing to the success of artificial intelligence (AI) in solving complicated problems, in this project we rely on the deep deterministic policy gradient (DDPG) technique, to manage the UAV assisted IoT resources. Our results show that DDPG achieves joint optimization of three objectives, namely sum data rate and harvested energy maximization, and energy consumption minimization, while out performing traditional mathematical schemes. The code of this project is made publicly accessible at https://github.com/QusaiBshiwa/Smart-6G- Sky-for-Green-Mobile-IOT-Networks