US 2026/0156022 A1Published application

Methods for enabling over-the-air federated learning using uplink pre-equalization

AssigneeInterDigital Patent Holdings, Inc.

InventorsAsil Koc, Philip Pietraski, Guodong Zhang, Mohamed Amine Arfaoui, John Kaewell

Life of this patent

Estimating the term…

Loading patent…

Abstract

A method implemented by a wireless transmit/receive unit (WTRU) may include receiving configuration information for over-the-air federated learning (OTA-FL) training rounds of an Artificial Intelligence/Machine Learning (AI/ML) model residing at the WTRU, which may include allocations of over-the-air reference signal (OTA-RS) resources and mechanisms for the uplink pre-equalization. An indication of uplink resource allocations for transmitting pre-equalized locally updated AI/ML model parameters may be received during the OTA-FL training rounds. A trained AI/ML model may be determined by local AI/ML model training. Local AI/ML model parameters or gradients may be generated for an FL task. Channel estimation may be performed during training rounds. Uplink pre-equalization parameters may be determined based on the channel estimation. Locally updated AI/ML model parameters may be pre-equalized for transmission. Pre-equalized locally updated AI/ML model parameters may be sent over uplink resources allocated during the OTA-FL training rounds. An updated global AI/ML model may be received.

Claims20

Dependent claims

A wireless transmit/receive unit (WTRU) comprising:

a processor configured to:
receive configuration information from a network for participating in one or more over-the-air federated learning (OTA-FL) training rounds of an Artificial Intelligence/Machine Learning (AI/ML) model residing at the WTRU, wherein the configuration information comprises an allocation of over-the-air reference signal (OTA-RS) resources and a mechanism to perform the uplink pre-equalization;
receive an indication of uplink resource allocations for transmitting pre-equalized locally updated AI/ML model parameters during the OTA-FL training rounds;
determine a trained AI/ML model through performing local AI/ML model training;
generate one or more of local AI/ML model parameters or gradients for a federated learning task;
perform channel estimation during the OTA-FL training rounds using OTA-RSs received via the OTA-RS resources;
determine uplink pre-equalization parameters based on the channel estimation;
pre-equalize the locally updated AI/ML model parameters for transmission on resource elements based on the uplink resource allocations and the pre-equalization parameters;
send the pre-equalized locally updated AI/ML model parameters during the OTA-FL training rounds over the uplink resources allocated according to the configuration information;
and
receive an updated global AI/ML model.

8 more elements

A method implemented by a wireless transmit/receive unit (WTRU), the method comprising:

receiving configuration information from a network for participating in one or more over-the-air federated learning (OTA-FL) training rounds of an Artificial Intelligence/Machine Learning (AI/ML) model residing at the WTRU, wherein the configuration information comprises an allocation of over-the-air reference signal (OTA-RS) resources and a mechanism to perform the uplink pre-equalization;
receiving an indication of uplink resource allocations for transmitting pre-equalized locally updated AI/ML model parameters during the OTA-FL training rounds;
determining a trained AI/ML model by performing local AI/ML model training;
generating one or more of local AI/ML model parameters or gradients for a federated learning task;
performing channel estimation during the OTA-FL training rounds using OTA-RSs received via the OTA-RS resources;
determining uplink pre-equalization parameters based on the channel estimation;
pre-equalizing the locally updated AI/ML model parameters for transmission on resource elements based on the uplink resource allocations and the pre-equalization parameters;
sending the pre-equalized locally updated AI/ML model parameters during the OTA-FL training rounds over the uplink resources allocated according to the configuration information;
and
receiving an updated global AI/ML model.

7 more elements

Dependencies are inferred from the claim text.