查看文章

Link adaptation is a promising tool of modern networks to combat the time-variant quality of channels. Modulation and Coding Scheme (MCS) selection is essentially used for link adaptation with channel dynamism. However, future generation networks need flexible link adaptation schemes that consider more parameters to improve the network performance. This paper proposes an energy-efficient link adaptation algorithm, in which a Deep Reinforcement Learning (DRL) agent is used to find the best match between the channel condition and the link parameters. Also, the downlink transmission power has been considered as a link parameter in addition to the modulation order and coding rate to make the link adaptation more flexible and efficient. Simulation results show that the proposed algorithm outperforms the benchmark algorithms regarding energy efficiency and link throughput.