A Deep-NN Beamforming Approach for Dual Function Radar-Communication THz UAV
Gianluca Fontanesi, Anna Guerra, Francesco Guidi, Juan A. Vásquez-Peralvo, Nir Shlezinger, Alberto Zanella, Eva Lagunas, Symeon Chatzinotas, Davide Dardari, Petar M. Djurić
Abstract
In this paper, we consider a scenario with one unmanned aerial vehicle (UAV), equipped with a uniform planar array (UPA), which transmits combined information and sensing signals to communicate with multiple ground base stations (GBSs) while simultaneously revealing the presence of potential targets within a specified area on the ground.We aim to jointly design the transmit beamforming and the GBSs association policyto optimize communication performance while ensuring high sensing accuracy. We propose a predictive beamforming framework based on a dual deep neural network (DNN) solution to solve the formulated nonconvex optimization problem. A first DNN is trained to generate the required transmit beamforming for any location within the UAV flying area more efficiently than traditional beamforming optimizer.A second DNN is trained to learn the optimal mapping from the input features, power, and effective isotropic radiated power (EIRP) constraints to the GBSs association decision. Finally, we provide an extensive simulation analysis to corroborate the proposed approach and show the benefits of EIRP, Signal-to-Noise-plus-Interference Ratio (SINR) performance and computational speed.