کنفرانس مهندسی برق ایران

صفحه اصلی / سی و چهارمین کنفرانس بین المللی مهندسی برق

GNN-Empowered Predictive Beamforming and UAV Deployment for Energy-Efficient aerial ISAC in Cluttered Vehicular Network

نویسندگان :

Sedighe Mirbolouk¹ Mohammad Amin Choukali²

1- دانشگاه ارومیه 2- دانشگاه ارومیه

کلمات کلیدی :

Graph neural network (GNN)،Integrated sensing and communication (ISAC)،Unmanned aerial vehicle (UAV)

چکیده :

UAV-enabled integrated sensing and communication (ISAC) has recently become a prominent research topic for sixth-generation (6G) wireless networks, where the unmanned aerial vehicles serve as aerial platforms to improve coverage and strengthen both sensing and communication capabilities. This paper proposes a joint predictive beamforming and UAV deployment scheme in a cluttered ISAC scenario to maximize the energy efficiency and meet the sensing Cramer-Rao lower bounds and vehicles’ quality of service (QoS). A graph learning based method is established to address the non-convexity of the considered optimization problem, where GNN and LSTM modules are employed to extract spatial and temporal dependencies of historical estimated channels. In this way, our method directly determines the UAV location and beamforming (BF) design for the subsequent time slot without any explicit channel state information. Finally, simulation results show the efficacy of the proposed method compared to other benchmarks.