WEIGHTED AVERAGE FEDERATED LEARNING BASED ON NEURAL NETWORK TRAINING LOSS: abstract simplified (17672533)

The abstract describes a method of wireless communication using a user equipment (UE) that involves updating an artificial neural network during a federated learning process. The updates can be in the form of gradients or updated model parameters. The method also involves keeping track of the training loss observed during the training of the neural network at each epoch of the federated learning process. Finally, the method includes transmitting the updates to a federated learning server, which will aggregate the gradients based on the training loss.