Abstract: A GAN training method, a machine learning system, and a communication apparatus are provided, which relate to the field of communication technologies and AI. The machine learning system includes a first device and a second device, where a first GAN runs on the first device, a second GAN runs on the second device, the first GAN includes a first generator and a first discriminator, and the second GAN includes a second generator and a second discriminator.

Abstract: A communication method and apparatus are provided. A modulation and coding scheme is determined based on a first norm of a first communication apparatus, and the first communication apparatus uploads information about a first learning model based on the modulation and coding scheme. The first norm of the first communication apparatus represents an importance degree of a local dataset of the first communication apparatus. The modulation and coding scheme used in a process in which a local training result is uploaded is determined based on the first norm, so that efficiency of federated learning can be improved.

Abstract: A data retransmission method, a relay station, a base station, and a communication system are provided. The data retransmission method includes: receiving a reception report sent by a base station and source data frames of a current transmission period that are sent by a user equipment through a subchannel, where the reception report carries sequence numbers of source data frames of a previous transmission period and the source data frames are not correctly received by the base station; obtaining according to the reception report and pre-stored source data frames of the previous transmission period, a set C1 of source data frames of the previous transmission period that need to be retransmitted; performing joint encoding for the source data frames in the set C1 and the source data frames of the current transmission period; and sending forwarded data frames generated by joint encoding to the base station through the subchannel.