TY - GEN
T1 - Digital Twin-Based Multi-UAV RIS Network for User Fairness and Network Load Balancing Using Multi-Tasking DRL
AU - Khan, Naveed
AU - Alwarafy, Abdulmalik
AU - Hayajneh, Mohamed
AU - Sallabi, Farag
AU - El-Sayed, Hesham
N1 - Publisher Copyright:
© 2025 IEEE.
PY - 2025
Y1 - 2025
N2 - Reconfigurable Intelligent Surfaces (RIS) have gained attention for enhancing wireless network performance, especially when integrated with Unmanned Aerial Vehicles (UAVs). This paper considers a multi-cell UAV and RIS-assisted HetNet with digital twin technology, focusing on network rate management to ensure user fairness(UF) and UAV-SC BSs load balancing(LB). Dual-connectivity UEs connect at the same time to mmWave UAV-SC BSs and sub-6GHz macrocells, with RIS-equipped UAVs enhancing mmWave reception and reducing interference. We formulate an optimization problem to jointly control active beamforming for UAV-SC-BSs and MC, as well as the passive beamforming and positions of RIS-equipped UAVs, aiming to maximize UF and LB. To address the complexity, we propose a Digital Twin Multi-Task Deep Reinforcement Learning (MTDRL) model based on the DDPG algorithm. Simulation results demonstrate that the proposed model achieves fair rate distribution and outperforms existing benchmarks.
AB - Reconfigurable Intelligent Surfaces (RIS) have gained attention for enhancing wireless network performance, especially when integrated with Unmanned Aerial Vehicles (UAVs). This paper considers a multi-cell UAV and RIS-assisted HetNet with digital twin technology, focusing on network rate management to ensure user fairness(UF) and UAV-SC BSs load balancing(LB). Dual-connectivity UEs connect at the same time to mmWave UAV-SC BSs and sub-6GHz macrocells, with RIS-equipped UAVs enhancing mmWave reception and reducing interference. We formulate an optimization problem to jointly control active beamforming for UAV-SC-BSs and MC, as well as the passive beamforming and positions of RIS-equipped UAVs, aiming to maximize UF and LB. To address the complexity, we propose a Digital Twin Multi-Task Deep Reinforcement Learning (MTDRL) model based on the DDPG algorithm. Simulation results demonstrate that the proposed model achieves fair rate distribution and outperforms existing benchmarks.
KW - Digital twin
KW - Heterogeneous Networks
KW - Load Balancing
KW - Multi-Task DRL
KW - Reconfigurable Intelligent Surface
KW - Unmanned Ariel Vehicles
KW - User Fairness
UR - https://www.scopus.com/pages/publications/105015659256
UR - https://www.scopus.com/pages/publications/105015659256#tab=citedBy
U2 - 10.1109/MeditCom64437.2025.11104471
DO - 10.1109/MeditCom64437.2025.11104471
M3 - Conference contribution
AN - SCOPUS:105015659256
T3 - 2025 IEEE International Mediterranean Conference on Communications and Networking, MeditCom 2025
SP - 24
EP - 29
BT - 2025 IEEE International Mediterranean Conference on Communications and Networking, MeditCom 2025
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2025 IEEE International Mediterranean Conference on Communications and Networking, MeditCom 2025
Y2 - 7 July 2025 through 10 July 2025
ER -