In this study, we deploy drones as dynamic base stations to address the issue of optimizing user coverage in areas without fixed base station infrastructure. To optimize drone placement, we employ Deep Q-Learning, beginning with a centralized approach due to its simplicity and ease of training. In this centralized approach, all drones are trained simultaneously. We also employ a decentralized technique in which each drone acts autonomously while sharing a common neural network, allowing for individualized learning. In addition, we explore the impacts of jamming on UAVs and provide a reliable approach for mitigating this interference. To boost robustness, we employ stochastic user distributions, which train our policy to successfully respond to a wide range of user situations.
In this work, we present an unmanned aerial vehicle (UAV) wireless dataset collected as part of the AERPAW Autonomous Aerial Data Mule (AADM) challenge, organized by the NSF Aerial Experimentation and Research Platform for Advanced Wireless (AERPAW) project. The AADM challenge was the second competition in which an autonomous UAV acted as a data mule, where the UAV downloaded data from multiple base stations (BSs) in a dynamic wireless environment. Participating teams designed flight control and decision-making algorithms for selecting which BSs to communicate with and how to plan flight trajectories to maximize data download within a mission completion time. The competition was conducted in two stages: Stage 1 involved development and experimentation using a digital twin (DT) environment, and in Stage 2, the final test run was conducted on the outdoor testbed. The total score for each team was compiled from both stages. The resulting dataset includes link quality and data download measurements, both in DT and physical environments. Along with the USRP measurements used in the contest, the dataset also includes UAV telemetry, Keysight RF sensor position estimates, link quality measurements from LoRa receivers, and Fortem radar measurements. It supports reproducible research on autonomous UAV networking, multi-cell association and scheduling, air-to-ground propagation modeling, DT-to-real-world transfer learning, and integrated sensing and communication, which serves as a benchmark for future autonomous wireless experimentation.
19 December 2025 in IEEE Open Journal of the Communications Society
Integrated sensing and communication (ISAC) is a promising technology for next-generation wireless networks, enabling high data rate transmission while providing precise sensing capabilities. However, the coexistence of sensing and communication functionalities can cause disruptive interference between these operations. Reconfigurable intelligent surfaces (RIS) provide a solution for efficient communications and high-precision sensing by intelligently controlling the reflection, phase, and direction of incident signals. In this paper, we formulate the optimization of base station (BS) beamforming and RIS phase shifts over the Riemannian manifold of Hermitian positive definite (HPD) matrices and the complex circle manifold respectively. We propose a model-driven meta-learning model in which recurrent networks perform alternating optimization (AO) by iteratively updating the beamforming matrices and RIS phase shifts through gradient steps on their respective manifolds. The model is trained using backpropagation through time (BPTT), enabling it to learn optimal initializations and adaptive gradient step sizes for both optimization variables. Simulation results demonstrate that the proposed manifold-based learning model achieves higher sum rates compared to its Euclidean-based counterpart.