"Improve UAV accuracy and loss testing by leveraging the Laplace equation and enhanced experience replay using least recently used (LRU) memory."

"The need to enhance coverage and capacity in response to unpredictable situations facing telecommunication operators has become a significant topic of conversation. Ground-based stations often fail to provide adequate coverage and capacity when a user transitions from one coverage area to another due to obstacles that limit signal strength. To address this recurrent issue, we utilize Unmanned Aerial Vehicles (UAV) as airborne base stations that receive directives from a controller network via the x2 interface to assist users in blind spots. The controller employs the Shapley Additive Explanation to choose the appropriate UAV for serving users, while the UAV applies the Laplace equation and Slack variables to attain optimal signal quality. The challenge is modeled as a Markov Decision Process, with Deep Deterministic Policy Gradient (DDPG) and Recurrent Twin Delayed Deep Deterministic Policy Gradient (RTD3) used to solve it. Furthermore, we employ the least recently used mechanism to enhance prioritized experience replay memory, thereby improving the training process of DDPG and RTD3 to address the UAV and controller problem, respectively. Extensive simulations have been conducted, and the results are superior to the baselines."