IP Library › Granted Patent US 11,646,838
Granted Patent B2
US 11,646,838 · App. 17/176,755 · Granted May 9, 2023

Broadcasting known data to train artificial neural networks

Inventors: Hamed Pezeshki (San Diego, CA); Tao Luo (San Diego, CA); Sony Akkarakaran (Poway, CA); Jingchao Bao (San Diego, CA); Ajay Gupta (San Diego, CA); Taesang Yoo (San Diego, CA); Vinod Viswanatha Menon (San Diego, CA); Wooseok Nam (San Diego, CA)
Assignee: QUALCOMM Incorporated
H04L5/0044H04L12/189H04W4/06H04W72/30H04W76/28
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 11,646,838
App. No.
17/176,755
Granted
May 9, 2023
Kind
B2
Abstract

A method of wireless communication, executed by a user equipment (UE), receives, from a base station, a broadcast or multicast message including a known payload, as well as a configuration for the known payload. The method also trains an artificial neural network with the known payload. A method of wireless communication, executed by a base station, configures a known payload for multiple UEs and signals, to the UEs, an indication of which physical channel will include the known payload, as well as time/frequency resources of the known payload. The method also broadcasts or multicasts the known payload to facilitate neural network training at the UEs.

Claims (44)

1. A method of wireless communication by a user equipment (UE), comprising:

receiving, from a base station, a configuration for a known payload, the known payload common to a plurality of UEs;

receiving, from the base station, a message including the known payload, the known payload comprising a ground truth label that is known to the base station and known to the UE without decoding of the message; and

performing online training of a machine learning model with the known payload.

2. The method of claim 1 , in which the configuration comprises a radio resource control (RRC) configured scrambling seed.

3. The method of claim 1 , further comprising receiving an indication of where to receive the known payload, the indication being either pre-configured or received via radio resource control (RRC) signaling, a media access control (MAC) control element or a physical downlink control channel (PDCCH).

4. The method of claim 3 , in which the indication defines a periodicity, duration, aggregation level, and/or payload size of the known payload.

5. The method of claim 1 , in which the known payload is located in a physical broadcast channel (PBCH), a physical downlink shared channel (PDSCH), or a physical downlink control channel (PDCCH).

6. The method of claim 1 , in which the message comprises a multicast message.

7. The method of claim 1 , in which the message comprises a broadcast message.

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

configuring a known payload for a plurality of user equipments (UEs);

signaling, to the plurality of UEs, an indication of which physical channel will include the known payload, and time/frequency resources of the known payload; and

transmitting, to the plurality of UEs, the known payload to enable the plurality of UEs to train machine learning models at the plurality of UEs, the known payload comprising a ground truth label that is known to the base station and known to the plurality of UEs without decoding the known payload.

9. The method of claim 8 , in which transmitting comprises multicasting the known payload.

10. The method of claim 8 , in which transmitting comprises broadcasting the known payload.

11. The method of claim 8 , in which the physical channel comprises a physical downlink control channel (PDCCH), a physical broadcast channel (PBCH), or a physical downlink shared channel (PDSCH) and the time/frequency resources comprise periodicity, duration, and payload size of the known payload.

12. The method of claim 8 , in which the known payload is defined within a physical downlink control channel (PDCCH).

13. The method of claim 8 , further comprising transmitting an message indicating where to receive the known payload, the message being either pre-configured or transmitted via radio resource control (RRC) signaling, a media access control (MAC) control element or a physical downlink control channel (PDCCH).

14. The method of claim 8 , in which configuring the known payload comprises transmitting a scrambling seed via radio resource control (RRC) signaling.

15. An apparatus of a user equipment (UE) for wireless communication, comprising:

a memory; and

at least one processor coupled to the memory and configured:

to receive, from a base station, a configuration for a known payload that is common to a plurality of UEs;

to receive, from the base station, a message including the known payload, the known payload comprising a ground truth label that is known to the base station and known to the UE without decoding of the message; and

to train a machine learning model with the known payload.

16. The apparatus of claim 15 , in which the configuration comprises a radio resource control (RRC) configured scrambling seed.

17. The apparatus of claim 15 , in which the at least one processor is further configured to receive an indication of where to receive the known payload, the indication being either pre-configured or received via radio resource control (RRC) signaling, a media access control (MAC) control element or a physical downlink control channel (PDCCH).

18. The apparatus of claim 17 , in which the indication defines a periodicity, duration, aggregation level and/or payload size of the known payload.

19. The apparatus of claim 15 , in which the known payload is located in a physical broadcast channel (PBCH), a physical downlink shared channel (PDSCH), or a physical downlink control channel (PDCCH).

20. The apparatus of claim 15 , in which the message comprises a multicast message.

21. The apparatus of claim 15 , in which the message comprises a broadcast message.

22. An apparatus of a base station for wireless communication, comprising:

a memory; and

at least one processor coupled to the memory and configured:

to configure a known payload for a plurality of user equipments (UEs);

to signal, to the plurality of UEs, an indication of which physical channel will include the known payload, and time/frequency resources of the known payload; and

to transmit, to the plurality of UEs, the known payload to enable the plurality of UEs to train machine learning models at the plurality of UEs, the known payload comprising a ground truth label that is known to the base station and known to the plurality of UEs without decoding the known payload.

23. The apparatus of claim 22 , in which the at least one processor is configured to transmit the known payload by multicasting the known payload.

24. The apparatus of claim 22 , in which the at least one processor is configured to transmit the known payload by broadcasting the known payload.

25. The apparatus of claim 22 , in which the physical channel comprises physical downlink control channel (PDCCH), physical broadcast channel (PBCH), or physical downlink shared channel (PDSCH) and the time/frequency resources comprise periodicity, duration, and payload size of the known payload.

26. The apparatus of claim 22 , in which the known payload is defined within a physical downlink control channel (PDCCH).

27. The apparatus of claim 22 , in which the at least one processor is further configured to send a message indicating where to receive the known payload, the message being either pre-configured or transmitted via radio resource control (RRC) signaling, a media access control (MAC) control element or a physical downlink control channel (PDCCH).

28. The apparatus of claim 22 , in which the at least one processor is further configured to configure the known payload by transmitting a scrambling seed via radio resource control (RRC) signaling.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Mar 16, 2021
From: PEZESHKI, HAMED; LUO, TAO; AKKARAKARAN, SONY; BAO, JINGCHAO; GUPTA, AJAY; YOO, TAESANG; MENON, VINOD VISWANATHA; NAM, WOOSEOK
To: QUALCOMM INCORPORATED
Reel/Frame 055605/0474 →
Continuity (2)
Provisional Application 62980070 · Feb 21, 2020
Related Publication 20210266125A1 · Aug 26, 2021
Cited By (1)
US 12,301,277