Novel multi-agent reinforcement learning for maximizing throughput in UAV-Enabled 5G networks

In beyond fifth-generation (B5G) networks, free-space optical (FSO) communication is anticipated to play a crucial role thanks to its high data rates and minimal system complexity. Therefore, infrequently occurring poor weather can impair its performance. The combination of FSO and radio frequency (RF) has proven a successful solution to address the increasing demand for high data rates in wireless communication networks. Due to their adaptability in terms of deployment and movement, unmanned aerial vehicles (UAVs) are also projected to be crucial in B5G networks. This paper investigates a UAV-aided hybrid FSO/DRF backhauling system using a matching game theory and Multi-Agent deep reinforcement learning (MARL) framework. We deploy a UAV to provide a user offloading service to an existing ground base station (GBS) facing a reduced backhaul capacity due to weather attenuation (e.g., fog). It is considered that the GBS has a pre-installed FSO backhaul connection to a macro-base station. However, during adverse weather conditions, the FSO backhaul is severely affected, compromising the reliability of the FSO link. The novelty here is the hybrid FSO/RF backhauling system and how it is utilized to address weather-related challenges. The UAV is deployed at the ideal height to increase system throughput using MARL, and the frequency division between the GBS and the UAV is also tuned. The system's effectiveness is assessed using meteorological data from the British cities of Edinburgh and London. The numerical outcomes demonstrate the suggested scheme's superiority and efficacy over traditional approaches. The time estimation shows that for a maximum of 30 users, the time consumed is 300 s which is lesser and effective.