Wireless LAN signal based motion detection method and device

Patent metadata

Field	Value
Publication number	US-12593304-B2
Application number	US-202318199366-A
Country	US
Kind code	B2
Filing date	May 19, 2023
Priority date	Dec 29, 2021
Publication date	Mar 31, 2026
Grant date	Mar 31, 2026

Abstract

Disclosed is a sensing device configured to perform movement detection by using a measurement value representing a channel state measured using a wireless LAN signal. The sensing device includes a memory, and a processor. The processor is configured to obtain a standard deviation of magnitudes of the measurement values over a predetermined number of measurements or within a predetermined time interval, and obtain a feature value of a sensing algorithm, based on the obtained standard deviation value. The sensing algorithm is an algorithm learned based on a correlation between a value obtained based on the magnitudes of the measurement values and the presence or absence of movement around devices exchanging the wireless LAN signal.

First claim

The invention claimed is:

1 . A sensing device configured to perform movement detection by using a measurement value representing a channel state measured using a wireless LAN signal, the sensing device comprising:
a memory; and a processor, wherein the processor is configured to: obtain a standard deviation of magnitudes of measurement values within a measurement window, wherein the measurement window is a time interval during which measurement is performed; and obtain a result of a movement detection of a sensing algorithm which is based on the obtained standard deviation value, and wherein the processor is further configured to use, as the measurement values, only a measurement value of an antenna having a smallest mean value of standard deviation of the measurement values within the measurement window, when the wireless LAN signal is simultaneously exchanged through a plurality of antennas of one terminal, and wherein the sensing algorithm is an algorithm learned based on a correlation between a value obtained based on the magnitudes of the measurement values and a presence or absence of movement around devices exchanging the wireless LAN signal.

2 . The sensing device of claim 1 , wherein the sensing algorithm is an algorithm learned based on a correlation between a value obtained based on the magnitudes of the measurement values and a strength of the wireless LAN signal and the presence or absence of movement around devices exchanging the wireless LAN signal.

3 . The sensing device of claim 1 , wherein the sensing algorithm is a binary classification function with respect to the movement detection,
the processor is configured to output whether a movement has been finally detected, based on whether a number of detection results indicating movement detection, among a plurality of detection results determined according to an output value of the sensing algorithm during a voting period, is equal to or greater than a predetermined number, and the voting period is a time interval determined by the sensing device.

4 . The sensing device of claim 3 , wherein the processor is configured to adjust a sensitivity of the movement detection by adjusting a length of the voting period.

5 . The sensing device of claim 1 , wherein the measurement value is obtained based on a sample mean, which is a value obtained by averaging magnitudes of channel state information (CSI) values measured at an identical time point in a plurality of subcarriers transmitting the wireless LAN signal.

6 . The sensing device of claim 5 , wherein the measurement value is a value filtered by a selective subcarrier shift removal filter,
the selective subcarrier shift removal filter is configured to apply Z-score normalization to samples corresponding to one or more outlier values, which are values outside a specific range, among sample mean values within the measurement window representing a time interval during which measurement is performed, and apply inverse Z-score normalization to the Z-score normalized sample by using a replacement mean and a replacement standard deviation, a minimum value in the specific range is a value obtained by subtracting a standard deviation of sample means within the specific range from a mean of sample means within the specific range, and a maximum value in the specific range is a value obtained by adding a standard deviation of sample means within the specific range to a mean of sample means within the specific range, the replacement mean is a mean value of sample mean values remaining after excluding the one or more outlier values from the sample mean values within the measurement window, the replacement standard deviation is a mean of sample standard deviations remaining after excluding the sample standard deviation for the one or more outlier values from sample standard deviation values, and the sample standard deviation is a standard deviation of magnitudes of CSI values measured at an identical time point in the plurality of subcarriers.

7 . The sensing device of claim 1 , wherein the measurement value is a value processed by an offset removal filter,
the offset removal filter is configured to obtain a difference between input values measured close together in time among a plurality of input values within the measurement window, apply an outlier removal filter to the obtained difference, and output a value obtained by adding a first input value which is temporally most advanced in the measurement window and difference values obtained by applying an outlier removal filtering to the first input value, the outlier removal filter is configured to replace an outlier value, which is a value outside a specific range among input values within the measurement window, with a replacement value, a minimum value in the specific range is a value obtained by subtracting a standard deviation of the input values within the specific range from a mean of the input values within the specific range, and the maximum value in the specific range is a value obtained by adding the standard deviation of input values within the specific range to the mean of input values within the specific range, and the replacement value is a mean value of input values remaining after excluding the outlier value from input values within the measurement window.

8 . A method for operating a sensing device configured to perform movement detection by using a measurement value representing a channel state measured using a wireless LAN signal, the method comprising:
obtaining a standard deviation of magnitudes of measurement values within a measurement window, wherein the measurement window is a time interval during which measurement is performed; and obtaining a result of a movement detection of a sensing algorithm; which is based on the obtained standard deviation value, wherein the obtaining the movement detection of the sensing algorithm comprises using, as the measurement value, only a measurement value of an antenna having a smallest mean value of standard deviation of the measurement values within the measurement window, when the wireless LAN signal is simultaneously exchanged through a plurality of antennas of one terminal, wherein the sensing algorithm is an algorithm learned based on a correlation between a value obtained based on the magnitudes of the measurement values and a presence or absence of movement around devices exchanging the wireless LAN signal.

9 . The method of claim 8 , wherein the sensing algorithm is an algorithm learned based on a correlation between a value obtained based on the magnitudes of the measurement values and a strength of the wireless LAN signal and the presence or absence of movement around devices exchanging the wireless LAN signal.

10 . The method of claim 8 , wherein the sensing algorithm is a binary classification function with respect to the movement detection,
the method further comprises outputting whether a movement has been finally detected, based on whether a number of detection results indicating movement detection, among a plurality of detection results determined according to an output value of the sensing algorithm during a voting period, is equal to or greater than a predetermined number, and the voting period is a time interval determined by the sensing device.

11 . The method of claim 10 , wherein the outputting whether a final movement has been finally detected comprises adjusting a sensitivity of the movement detection by adjusting a length of the voting period.

12 . The method of claim 8 , wherein the measurement value is obtained based on a sample mean, which is a value obtained by averaging magnitudes of channel state i

Assignees

• Wilus Inst Standards & Tech Inc

Inventors

• Zebulon Asaf
• Pranaya Fathan Adi
• Hwang Hwanwoong
• Son Juhyung
• Kwak Jinsam

Classifications

• G01S5/0294

  Trajectory determination or predictive filtering, e.g. target tracking or Kalman filtering · CPC title

• G01S5/021

  Calibration, monitoring or correction (G01S5/0252 takes precedence) · CPC title

• H04W84/12

  WLAN [Wireless Local Area Networks] · CPC title

• H04B17/309

  Measuring or estimating channel quality parameters · CPC title

• G01S5/017

  Detecting state or type of motion · CPC title