Intelligent Spectrum Allocation for TN-NTN Integration: A Q-Learning Approach

Sixth generation (6G) wireless networks connect people all over the world by combining terrestrial networks (TN) with non-terrestrial networks (NTN), such as satellites and aerial platforms. It is hard to allocate spectrum because of changes in the topology, interference, and quality-of-service issues. This study presents a self-directed channel assignment system utilizing Q-learning. The results of 5,000 training episodes on 10 channels with 15 TN and 8 NTN consumers show convergence, with an average reward of 40.37±12.01 and a throughput of 40.49 Mbps. The policy is fair and changes based on traffic patterns with Jain's index of 0.75.