Granted Patent B2

US 12,537,718 · App. 18/386,894 · Granted Jan 27, 2026

Inventor: Eunsung Jeon (Suwon-si, KR)

Assignee: SAMSUNG ELECTRONICS CO., LTD.

H04L25/0202H04B7/043H04B17/336

Abstract

Claims (47)

1 . An operating method of a first apparatus communicating with a second apparatus through a wireless local area network (WLAN), the operating method comprising:

obtaining channel characteristic data with the second apparatus based on a null data packet (NDP) frame received from the second apparatus;

generating beamforming feedback information and a feedback frame based on a class of a channel determined by applying a machine learning algorithm to the channel characteristic data; and

transmitting the feedback frame to the second apparatus.

2 . The operating method of claim 1 , wherein the channel characteristic data comprises at least one of a variance of a channel frequency response of the channel, a channel delay spread of the channel, or a signal-to-noise ratio (SNR) of the channel.

3 . The operating method of claim 1 , wherein the class of the channel comprises a first channel class and a second channel class, and

wherein a second multipath fading of the second channel class is greater than a first multipath fading of the first channel class.

4 . The operating method of claim 1 , wherein the obtaining the channel characteristic data comprises:

identifying the NDP frame;

extracting pieces of information included in the NDP frame; and

obtaining the channel characteristic data of the second apparatus based on the extracted pieces of information.

5 . The operating method of claim 1 , wherein the generating the beamforming feedback information and the feedback frame comprises:

obtaining pieces of channel training data comprising first pieces of first channel class data and second pieces of second channel class data;

determining K pieces of K-nearest neighbor data closest to the channel characteristic data from among the pieces of channel training data, K being a positive integer greater than zero;

comparing a first number of pieces of the first channel class data from among the K pieces of K-nearest neighbor data with a second number of pieces of second channel class data from among the K pieces of K-nearest neighbor data; and

determining, based on the comparing, the class of the channel corresponding to the channel characteristic data to be the class of the channel corresponding to a greater number between the first number of pieces and the second number of pieces.

6 . The operating method of claim 1 , wherein the generating the beamforming feedback information and the feedback frame comprises determining the class of the channel by applying an artificial neural network algorithm to the channel characteristic data.

7 . The operating method of claim 1 , wherein the beamforming feedback information comprises subcarrier grouping information corresponding to a number of subcarriers grouped from among consecutive subcarriers.

8 . The operating method of claim 1 , wherein the feedback frame comprises pieces of quantization angle information corresponding to the channel characteristic data, and

wherein the beamforming feedback information comprises codebook size information corresponding to a number of bits of the pieces of quantization angle information.

9 . An operating method of a second apparatus communicating with a first apparatus through a wireless local area network (WLAN), the operating method comprising:

transmitting a null data packet (NDP) frame to the first apparatus;

receiving a channel information frame from the first apparatus;

obtaining channel characteristic data with the first apparatus based on the channel information frame;

generating beamforming feedback information based on a class of a channel determined by applying a machine learning algorithm to the channel characteristic data; and

transmitting the beamforming feedback information to the first apparatus.

10 . The operating method of claim 9 , wherein the channel characteristic data comprises at least one of a variance of a channel frequency response of the channel, a channel delay spread of the channel, or a signal-to-noise ratio (SNR) of the channel.

11 . The operating method of claim 9 , wherein the class of the channel comprises a first channel class and a second channel class, and

wherein a second multipath fading of the second channel class is greater than a first multipath fading of the first channel class.

12 . The operating method of claim 9 , wherein the generating the beamforming feedback information comprises:

obtaining pieces of first channel training data comprising pieces of initial center point data;

determining nearest center point data closest to the channel characteristic data; and

determining the class of the channel corresponding to the channel characteristic data to be the class of the channel corresponding to the nearest center point data.

13 . The operating method of claim 12 , wherein the generating the beamforming feedback information further comprises updating pieces of center point data based on the pieces of first channel training data and pieces of second channel training data comprising the channel characteristic data.

14 . The operating method of claim 9 , wherein the beamforming feedback information comprises subcarrier grouping information corresponding to a number of subcarriers grouped from among consecutive subcarriers.

15 . The operating method of claim 9 , further comprising:

receiving, from the first apparatus, a feedback frame comprising pieces of quantization angle information corresponding to the channel characteristic data,

wherein the beamforming feedback information comprises codebook size information corresponding to a number of bits of the pieces of quantization angle information.

16 . The operating method of claim 9 , wherein the WLAN is a multi-user multi-input multi-output (MU-MIMO) system.

17 . A first apparatus communicating with a second apparatus through a wireless local area network (WLAN), the first apparatus comprising:

a processing circuit configured to obtain channel characteristic data based on a null data packet (NDP) frame received from the second apparatus, and generate beamforming feedback information and a feedback frame based on a class of a channel determined by applying a machine learning algorithm to the channel characteristic data,

wherein the processing circuit is further configured to control a transceiver to transmit the feedback frame to the second apparatus.

18 . The first apparatus of claim 17 , wherein the channel characteristic data comprises at least one of a variance of a channel frequency response of the channel, a channel delay spread of the channel, or a signal-to-noise ratio (SNR) of the channel.

19 . The first apparatus of claim 17 , wherein the class of the channel comprises a first channel class and a second channel class, and

wherein a second multipath fading of the second channel class is greater than a first multipath fading of the first channel class.

20 . The first apparatus of claim 17 , wherein the feedback frame comprises pieces of quantization angle information corresponding to the channel characteristic data, and

wherein the beamforming feedback information comprises codebook size information corresponding to a number of bits of the pieces of quantization angle information.

Assignments (1)

Priority Claims (3)

Continuity (1)

References Cited (26)