Conclusions

This chapter concludes the Brief. Since mobile data traffic has kept increasing in recent years, spectrum sharing has emerged as a promising way to alleviate the contradiction between the limited spectrum resource and the increasing mobile data traffic in wireless edge networks. In this Brief, we investigate the AI-enabled spectrum sharing in wireless edge networks and focus on different spectrum sharing scenarios. Specifically, we leverage reinforcement learning to enhance communication performance through several proposed methods, i.e., DQN-based MCS selection method for opportunistic scenarios, DDPG-based power control method for centralized scenarios and DDPG-based power control method for distributed scenarios. In particular, our proposed methods could be further studied in more practical situations by considering the convergence time in various scenarios and the synchronization designs, and has the potential to be extended into space-air-ground integrated networks in the future.