Cooperative Deep Learning-Based Uplink Distributed Fair Resource Allocation for Aerial Reconfigurable Intelligent Surfaces Wireless Networks

In this paper, we present a cooperative distributed learning based resource allocation scheme for an unmanned aerial vehicle (UAV)-assisted wireless communication framework equipped with reconfigurable intelligent surfaces (RIS), aiming to dynamically allocate the amount of the resource block(RB) to ensure the fairness and efficiency from the end-to-end communications system perspective. We suggest a novel, but simple fairness-efficiency metrics (FE metrics) to measure the performance of resource allocating, and to compare various resource allocation policies, including our proposed cooperative deep learning approaches, other rule-based and centralized methods. The proposed cooperative distributed learning framework has the advantages of making decisions by considering information about the entire environment, so that cooperation provides better overall performance and reliability. Moreover, it enables double parallel computing and reduces execution time dramatically. The simulation result shows that our cooperative deep learning algorithm ensures excellent convergence and low time complexity during the execution in the complex Aerial RIS system to reduce communication delays, improve communication quality, and ensure fairness.