Federated Dueling DQN based Joint Beamformng and Power Alloction in UAV Network

To address the problem of communication quality degradation due to inter-link interference in unmanned aerial vehicle (UAV) network, a joint beamforming and power allocation (JBPA) optimization model is developed to maximize the spectrum efficiency (SE) and energy efficiency (EE) of UAV systems under the premise of ensuring the quality of service (QoS) from the space and power domains. Then, a JBPA strategy based on federated dueling DQN (F-DDQN) learning algorithm, which is an effective technology to handle communication costs and data privacy, is designed to solve the presented optimization model. Simulation results show that the proposed F-DDQN based JBPA strategy improves the spectrum efficiency and energy efficiency of UAV network, and has the advantage of rapid convergence.