Federated learning with varying feedback

A method of wireless communication, by a base station, comprising:

• assigning to each of a plurality of user equipments (UEs), a number and/or a location of digits of a value for transmission based on a deterministic task assignment rule or a probabilistic task assignment rule, the value representing at least one of (1) a gradient estimate of a weight of a jointly trained artificial neural network or (2) the weight of the neural network;
• transmitting the deterministic task assignment rule or the probabilistic task assignment rule to the plurality of UEs;
• receiving, from each of the plurality of UEs, the number and/or location of a plurality of digits of the value, in accordance with the assigning;
• and
• training the neural network based on the plurality of digits received from the plurality of UEs.

A user equipment (UE), comprising:

• a processor;
• a memory coupled with the processor;
• and
• instructions stored in the memory and operable, when executed by the processor, to cause the UE:
• to receive, from a base station, a jointly trained artificial neural network;
• to calculate a value representing at least one of (1) a gradient estimate for a weight of the jointly trained artificial neural network, or (2) the weight of the jointly trained artificial neural network;
• to expand the value into a numerical system with base N into a plurality of digits;
• to determine a number and/or a location of the plurality of digits to transmit based on a deterministic task assignment rule received from the base station or a probabilistic task assignment rule;
• and
• to transmit the determined number and/or the determined location of the plurality of digits to the base station.

A base station, comprising:

• a processor;
• a memory coupled with the processor;
• and
• instructions stored in the memory and operable, when executed by the processor, to cause the base station:
• to assign to each of a plurality of user equipments (UEs), a number and/or a location of digits of a value for transmission based on a deterministic task assignment rule or a probabilistic task assignment rule, the value representing at least one of (1) a gradient estimate of a weight of a jointly trained artificial neural network or (2) the weight of the neural network;
• to transmit the deterministic task assignment rule or the probabilistic task assignment rule to the plurality of UEs;
• to receive, from each of the plurality of UEs, the number and/or location of a plurality of digits of the value, in accordance with the assigning;
• and
• to train the neural network based on the plurality of digits received from the plurality of UEs.