Deep reinforcement learning based secondary user transmit power control for underlay cognitive radio networks
Kouhei Katou, Xiaoyan Wang, Masahiro Umehira, Yusheng Ji · 2022
To improve the spectral utilization efficiency, underlay cognitive radio network has been extensively investigated in recent years. The underlay paradigm allows secondary users to operate if the interference they cause to primary user is below a given threshold. The transmit power control problem of the secondary user is critical and challenging for underlay cognitive radio networks, especially when the network scenario is dynamic. In this work, we propose a deep reinforcement learning based secondary user transmit power control scheme for underlay cognitive radio networks. The proposed scheme dynamically controls the transmit power of a mobile secondary user, with the purpose of improving the system's spectral utilization efficiency and meanwhile satisfying the required SINR (signal-to-interference plus noise power ratio) for primary users. The performance of the proposed scheme in terms of interference ratio and throughput is validated by extensive simulations.