公开(公告)号：US20230308920A1

公开(公告)日：2023-09-28

申请号：US18041090

申请日：2021-06-25

Applicant: QUALCOMM Incorporated

Inventor： Alexandros MANOLAKOS ,  June NAMGOONG ,  Taesang YOO ,  Naga BHUSHAN ,  Pavan Kumar VITTHALADEVUNI ,  Jay Kumar SUNDARARAJAN ,  Krishna Kiran MUKKAVILLI ,  Hwan Joon KWON ,  Tingfang JI

IPC: H04W24/10 ,  G06N3/0455 ,  G06N3/098

CPC classification number: H04W24/10 ,  G06N3/0455 ,  G06N3/098

Abstract: Various aspects of the present disclosure relate to wireless communication. In some aspects, a client may receive a selection feedback configuration associated with a reporting procedure for reporting updates corresponding to at least one autoencoder index associated with one or more autoencoders selected by the client. The client may determine an update corresponding to the at least one autoencoder index. The client may transmit the update based at least in part on the selection feedback configuration. Numerous other aspects are provided.

公开(公告)号：US20230274194A1

公开(公告)日：2023-08-31

申请号：US17584332

申请日：2022-01-25

Applicant: QUALCOMM Incorporated

Inventor： Srinivas YERRAMALLI ,  Taesang YOO ,  Rajat PRAKASH ,  Junyi LI ,  Eren BALEVI ,  Hamed PEZESHKI ,  Tao LUO ,  Xiaoxia ZHANG ,  Aziz Gholmieh

IPC: G06N20/20 ,  H04L1/18

CPC classification number: G06N20/20 ,  H04L1/1835

Abstract: A method of wireless communication by a user equipment (UE) includes receiving, from a network device, a request to initiate gradient computations for a round of federated learning. The method also includes computing gradients in response to receiving the request to initiate gradient computation. The method further includes informing the network device of availability of the gradients. The method still further includes receiving, from the network device, information to enable transfer of the gradients to the network device. The method further includes transferring the gradients to the network device in response to receiving the information to enable transfer.

公开(公告)号：US20230262483A1

公开(公告)日：2023-08-17

申请号：US17584331

申请日：2022-01-25

Applicant: QUALCOMM Incorporated

Inventor： Srinivas YERRAMALLI ,  Taesang YOO ,  Rajat PRAKASH ,  Junyi LI ,  Eren BALEVI ,  Hamed PEZESHKI ,  Tao LUO ,  Xiaoxia ZHANG ,  Aziz GHOLMIEH

IPC: H04W24/02 ,  H04L41/16

CPC classification number: H04W24/02 ,  H04L41/16

Abstract: A protocol stack architecture for processing machine learning (ML) data includes a ML layer to manage ML data communication with a network device. The ML layer is coupled to multiple ML training blocks, and ML and inference blocks for multiple neural networks, and an analog data communications stack coupled to the ML layer. The analog data communications stack has an upper media access control analog (MAC-A) layer coupled to the ML layer and configured to store data for each neural network, a lower MAC-A layer coupled to the upper MAC-A layer and configured to segment and reassemble analog ML data, and an analog physical layer coupled to the lower MAC-A layer and configured to communicate analog data with the network device. The architecture includes a digital data communications stack coupled to the ML layer and the lower MAC-A layer and configured to manage digital communications with the network device.

公开(公告)号：US20230259754A1

公开(公告)日：2023-08-17

申请号：US17585399

申请日：2022-01-26

Applicant: QUALCOMM Incorporated

Inventor： Xipeng ZHU ,  Taesang YOO ,  Gavin Bernard HORN

IPC: G06N3/08

CPC classification number: G06N3/08 ,  H04W16/14

Abstract: A method of wireless communication by a first network device includes receiving a machine learning model, the machine learning model being unvalidated and trained. The method also includes quantizing and compiling the machine learning model. The method further includes obtaining verification data. The method still further includes validating the machine learning model with the verification data to determine a performance level of the machine learning model. A method of wireless communication by a first network device includes transmitting machine learning model verification data to a second network device. The method also includes receiving a report, from the second network device, indicating performance of a machine learning model with the verification data. The method further includes validating the machine learning model in response to the performance satisfying a performance condition.

公开(公告)号：US20230254773A1

公开(公告)日：2023-08-10

申请号：US18003059

申请日：2021-08-13

Applicant: QUALCOMM Incorporated

Inventor： Pavan Kumar VITTHALADEVUNI ,  Alexandros MANOLAKOS ,  Taesang YOO ,  Naga BHUSHAN ,  June NAMGOONG ,  Jay Kumar SUNDARARAJAN ,  Krishna Kiran MUKKAVILLI ,  Wanshi CHEN ,  Tingfang JI

IPC: H04W52/02 ,  H04B7/06

CPC classification number: H04W52/0261 ,  H04B7/0626 ,  H04B7/0641

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a first device may determine that a power threshold for the first device is satisfied. The first device may transition from a first type of channel state feedback processing to a second type of channel state feedback processing based at least in part on determining that the power threshold for the first device is satisfied.

公开(公告)号：US20230246693A1

公开(公告)日：2023-08-03

申请号：US18004611

申请日：2021-08-13

Applicant: QUALCOMM Incorporated

Inventor： Pavan Kumar VITTHALADEVUNI ,  Hwan Joon KWON ,  Taesang YOO ,  Alexandros MANOLAKOS ,  Naga BHUSHAN ,  June NAMGOONG ,  Jay Kumar SUNDARARAJAN ,  Krishna Kiran MUKKAVILLI ,  Tingfang JI

IPC: H04B7/06 ,  H04L1/00

CPC classification number: H04B7/0641 ,  H04L1/0026 ,  H04B7/0626 ,  H04L1/0029

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a first device may receive a channel state information (CSI) feedback configuration comprising an indication to save, for a specified time period, a channel state feedback (CSF) that corresponds to a first reference signal carried on a downlink channel. The first device may transmit a differential CSF based at least in part on the CSF and a second reference signal carried on the downlink channel. Numerous other aspects are provided.

公开(公告)号：US20230239239A1

公开(公告)日：2023-07-27

申请号：US17584345

申请日：2022-01-25

Applicant: QUALCOMM Incorporated

Inventor： Srinivas YERRAMALLI ,  Taesang YOO ,  Rajat PRAKASH ,  Junyi LI ,  Eren BALEVI ,  Hamed PEZESHKI ,  Tao LUO ,  Xiaoxia ZHANG ,  Aziz Gholmieh

IPC: H04L45/00 ,  H04L45/52 ,  H04W40/02 ,  G06N20/20

CPC classification number: H04L45/66 ,  G06N20/20 ,  H04L45/52 ,  H04W40/02

Abstract: A method of wireless communication by a user equipment (UE) includes generating, by an upper analog media access control (MAC-A) layer of a protocol stack, a data packet with a header and a data field. The header indicates a neural network identifier (ID) and a request ID. The data field includes gradient data for a federated learning iteration. The method also includes transferring the data packet to lower layers of the protocol stack for transmission to a network device across a wireless network.

公开(公告)号：US20230232361A1

公开(公告)日：2023-07-20

申请号：US17648239

申请日：2022-01-18

Applicant: QUALCOMM Incorporated

Inventor： Mohammed Ali Mohammed HIRZALLAH ,  Srinivas YERRAMALLI ,  Rajat PRAKASH ,  Taesang YOO ,  Xiaoxia ZHANG ,  Roohollah AMIRI ,  Marwen ZORGUI

IPC: H04W64/00 ,  H04W72/04

CPC classification number: H04W64/00 ,  H04W72/0406 ,  H04W72/048

Abstract: Disclosed are techniques for wireless positioning. In an aspect, a first user equipment (UE) obtains one or more first radio frequency fingerprint (RFFP) measurements of one or more first downlink channels received at the first UE, one or more first sidelink channels received at the first UE, or both, and determines one or more locations of a target UE based on the one or more first RFFP measurements and a machine learning module, wherein the machine learning module is trained based on previously collected RFFP measurements of one or more downlink channels, RFFP measurements of one or more uplink channels, RFFP measurements of one or more sidelink channels, locations of one or more sidelink anchor UEs, locations of one or more base stations, or any combination thereof.

公开(公告)号：US20230188302A1

公开(公告)日：2023-06-15

申请号：US18004286

申请日：2020-08-31

Applicant: QUALCOMM INCORPORATED

Inventor： Pavan Kumar VITTHALADEVUNI ,  Taesang YOO ,  Naga BHUSHAN ,  June NAMGOONG ,  Bo CHEN ,  Ruifeng MA ,  Krishna Kiran MUKKAVILLI ,  Tingfang JI

IPC: H04L5/00 ,  H04W72/231 ,  G06N3/08

CPC classification number: H04L5/0057 ,  H04L5/0048 ,  H04W72/231 ,  G06N3/08

Abstract: Methods, systems, and devices for wireless communications are described. Generally, the described techniques at a user equipment (UE) provide for efficiently reporting channel state information (CSI) to a base station with an appropriate level of accuracy. In particular, the base station may indicate a level of accuracy to the UE for reporting CSI. The UE may encode the CSI using a first neural network, and the base station may decode the CSI using a second neural network. The first and second neural networks may form a neural network pair, and the UE may train the neural network pair based on the level of accuracy indicated by the base station. For example, the base station may indicate a loss function corresponding to a level of accuracy with which CSI is to be reported by the UE, and the UE may train the neural network pair using the loss function.

公开(公告)号：US20230179953A1

公开(公告)日：2023-06-08

申请号：US17457718

申请日：2021-12-06

Applicant: QUALCOMM Incorporated

Inventor： Mohammed Ali Mohammed HIRZALLAH ,  Srinivas YERRAMALLI ,  Taesang YOO ,  Rajat PRAKASH ,  Xiaoxia ZHANG

IPC: H04W4/029 ,  H04B7/06 ,  H04B17/318

CPC classification number: H04W4/029 ,  H04B7/0617 ,  H04B17/318

Abstract: Aspects presented herein may enable an ML module to associate RF fingerprints with beam directions and/or beam features to improve the uniqueness of RF fingerprints. In one aspect, network entity may receive, from one or more wireless devices, a plurality of first RF fingerprints, each of the plurality of first RF fingerprints being associated with at least one directional feature and a location. The network entity may receive a request to determine a position of a UE based on at least one second RF fingerprint associated with the UE or captured by the UE. The network entity may estimate the position of the UE based at least in part on matching the at least one second RF fingerprint to at least one of the plurality of first RF fingerprints.