Social-Aware Peer Selection for Energy Efficient D2D Communications in UAV-Assisted Networks: A Q-Learning Approach
Aamir Nadeem, Arif Ullah, Wooyeol Choi
Abstract
Leveraging the benefits of both technologies, unmanned aerial vehicles (UAV) and device-to-device (D2D) communications can be jointly utilized to assist each other and improve network performance. However, best peer selection in such networks is a challenging task. In this letter, we propose a UAV-assisted scheme that leverages the UAVs’ air-to-ground channel to improve peer selection in social-aware D2D networks. Furthermore, a Q-learning algorithm has been employed to learn the policy for selecting the best D2D peers based on users’ social and physical parameters. Extensive simulations demonstrate that the proposed Q-learning-based peer selection outperforms the existing TOPSIS-based and random peer selection algorithms in terms of average data rate, and energy efficiency.