Deep Reinforcement Learning based Usage Aware Spectrum Access Scheme
Yuto Teraki, Xiaoyan Wang, Masahiro Umehira, Yusheng Ji · 2021
To deal with the spectrum-shortage problem, dynamic spectrum access (DSA) has attracted a great deal of attention in both academia and industry. In DSA, secondary users (SUs) are allowed to exploit the whitespace of the primary users (PUs) on an instant-by-instant basis. The goal is to improve the system’s spectral utilization efficiency in a manner that limits the interference from SUs to PUs. To this end, in this paper, we proposed an usage aware spectrum access scheme by exploiting deep reinforcement learning. We evaluated its performance by extensive simulations, and validate the superiority of the proposed scheme by comparing it with existing methods.