As the demand for spectrum increases, the coexistence of terrestrial networks (TNs) and non-terrestrial networks (NTNs) is becoming a critical aspect of 6th generation (6G) communication scenarios. Integrated TN-NTN systems share the same frequency bands, leading to significant performance degradation due to co-channel interference. This paper proposes a deep Q-network (DQN) based deep reinforcement learning (DRL) algorithm to efficiently allocate resource blocks within shared frequency bands to each user equipment (UE) in integrated TN-NTN systems. The proposed algorithm continuously interacts with the network environment to learn optimal resource allocation policies. The proposed algorithm continuously interacts with the network environment to learn optimal resource allocation policies, thereby minimizing interference and maximizing system data throughput, promoting efficient use of spectrum resources.
This work was supported by the Institute of Information and Communications Technology Planning and Evaluation (IITP) grant funded by Korean Government the Ministry of Science and Information and Communications Technology (MSIT), Information and Communications Technology (Development of 3D Spatial Satellite Communications Technology) under Grant 2021-0-00847
