Q-Learning NOMA Random Access for IoT-Satellite Terrestrial Relay Networks
Douglas Alisson Tubiana, Jamil Farhat, Glauber Brante, Richard Demo Souza
Abstract
We consider an IoT-satellite terrestrial relay network providing access for a massive quantity of IoT devices. In order to overcome the difficulties associated with random access over ultra-dense IoT networks, a <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${Q}$ </tex-math></inline-formula> -Learning non-orthogonal multiple access (NOMA)-based random access protocol is designed. The proposed QL-NOMA algorithm allocates timeslots and communication channels in the link between IoT devices and the relays, aiming at improved throughput. Numerical results show that the proposed method can significantly boost the network performance in comparison with Slotted-ALOHA with NOMA, while being a decentralized strategy with minimal feedback. Additionally, our results show that there is an optimal quantity of relay channels depending on the number of IoT devices trying to access the network.