Granted Patent B2

US 12,418,784 · App. 17/921,315 · Granted Sep 16, 2025

Inventors: Hongwon Lee (Seoul, KR); Hangyu Cho (Seoul, KR); Jeonghwan Yoon (Seoul, KR); Homin Yoo (Seoul, KR); Taesung Lim (Seoul, KR)

Assignee: LG ELECTRONICS INC.

H04W8/005G06N3/088G06N3/09H04W8/24H04W84/12

Abstract

Claims (63)

1. A method by a user identification device in a wireless local area network (WLAN) system, the method comprising:

discovering a PED (Personal Electric Device) and obtaining identification information of the PED;

collecting data and pre-processing the collected data based on the identification information of the PED; and

generating a user identification model by learning the pre-processed data,

wherein the pre-processed data is data in which the collected data and identification information of the PED are matched,

wherein the user identification model is generated based on supervised learning, and

wherein the supervised learning uses the identification information of the PED as a correct answer.

2. The method of claim 1 , wherein the PED is discovered through Bluetooth, Wi-Fi, Zigbee, or Near Field Communication (NFC).

3. The method of claim 1 , wherein the identification information of the PED is obtained from the PED or a cloud server in which the PED is registered,

wherein the identification information of the PED includes a person's name, an International Mobile Equipment Identity (IMEI), a MAC address, or a serial number.

4. The method of claim 1 , wherein the pre-processed data includes first and second data,

wherein the first data is data in which the collected data and identification information of the PED are matched,

wherein the second data is collected data that does not match the identification information of the PED,

wherein the user identification model is generated based on semi-supervised learning,

wherein the semi-supervised learning uses the identification information of the PED as a correct answer.

5. The method of claim 1 , further comprising:

generating a cluster based on unsupervised learning for the collected data; and

matching identification information of the PED to the cluster,

wherein the cluster matching the identification information of the PED includes classified data by category.

6. The method of claim 5 , wherein the user identification model is generated based on supervised learning for the data classified by category,

wherein the supervised learning uses the identification information of the PED as a correct answer.

7. The method of claim 1 , wherein when the user identification model is initially generated, further comprising:

generating a result value based on the initially generated user identification model and identification information of the PED; and

generating a modified user identification model by re-learning on the result value.

8. The method of claim 1 , further comprising:

identifying a user of a wireless signal based on the user identification model.

9. The method of claim 1 ,

wherein the PED includes an identification information providing part, a user identification device discovery part, and an identification information exchange network part,

wherein the user identification device includes an identification information exchange network part, a PED discovery part, a wireless sensing signal collection part, a wireless sensing signal processing part, a user identification learning part, and a user identification prediction part,

wherein the PED is discovered through the PED discovery part,

wherein the data is collected through the wireless sensing signal collection part,

wherein the collected data is pre-processed through the wireless sensing signal processing part,

wherein the pre-processed data is learned through the user identification learning part, and

wherein the user identification prediction part predicts a user through the generated user identification model.

10. A user identification device in a wireless local area network (WLAN) system, the user identification device comprising:

a memory;

a transceiver; and

a processor being operatively connected to the memory and the transceiver,

wherein the processor is configured to:

discover a PED (Personal Electric Device) and obtaining identification information of the PED;

collect data and pre-processing the collected data based on the identification information of the PED; and

generate a user identification model by learning the pre-processed data,

wherein the pre-processed data is data in which the collected data and identification information of the PED are matched,

wherein the user identification model is generated based on supervised learning, and

wherein the supervised learning uses the identification information of the PED as a correct answer.

11. The user identification device of claim 10 , wherein the identification information of the PED is obtained from the PED or a cloud server in which the PED is registered,

wherein the identification information of the PED includes a person's name, an International Mobile Equipment Identity (IMEI), a MAC address, or a serial number.

12. The user identification device of claim 10 , wherein the pre-processed data includes first and second data,

wherein the first data is data in which the collected data and identification information of the PED are matched,

wherein the second data is collected data that does not match the identification information of the PED,

wherein the user identification model is generated based on semi-supervised learning,

wherein the semi-supervised learning uses the identification information of the PED as a correct answer.

13. The user identification device of claim 10 , wherein the processor is further configured to:

generate a cluster based on unsupervised learning for the collected data; and

match identification information of the PED to the cluster,

wherein the cluster matching the identification information of the PED includes classified data by category.

14. The user identification device of claim 13 , wherein the user identification model is generated based on supervised learning for the data classified by category,

wherein the supervised learning uses the identification information of the PED as a correct answer.

15. The user identification device of claim 10 , wherein when the user identification model is initially generated, wherein the processor is further configured to:

generate a result value based on the initially generated user identification model and identification information of the PED; and

generate a modified user identification model by re-learning on the result value.

16. The user identification device of claim 10 , wherein the processor is further configured to:

identify a user of a wireless signal based on the user identification model.

Assignments (1)

Continuity (1)

References Cited (9)