Method and device for motion sensing based on wireless LAN signal

Abstract

Disclosed is a sensing device for sensing respiration of a subject by using a measurement value indicating a channel state measured using a wireless LAN signal. The sensing device comprises a memory and a processor. The processor acquires a measurement value by measuring a channel state on the basis of a wireless LAN signal received by the sensing device from a sensing node, and senses respiration of a subject within a sensing radius on the basis of a cross-correlation operation between a pre-designated basic signal and a target signal in which the measurement value is arranged in time series.

Description

Method and device for motion detection based on wireless LAN signals

[0001] The present invention relates to a method and device for detecting motion using a wireless LAN signal.

[0002] With the recent proliferation of mobile devices, wireless LAN (WLAN) technology, which can provide them with fast wireless Internet service, is attracting significant attention. WLAN technology utilizes short-range wireless communication technology to enable mobile devices such as smartphones, tablets, laptops, portable multimedia players, and embedded devices to wirelessly connect to the Internet at home, in businesses, or in specific service areas.

[0003] With the proliferation of wireless LAN, research is also underway to detect motion around access points (APs) and stations using the signals exchanged between them. Motion detection using WLAN signals has the advantage of minimizing unnecessary exposure of privacy, as it eliminates the need for direct recording of the user's appearance. However, further research is needed to improve the accuracy of motion detection and to avoid interfering with existing WLAN communications.

[0004] An embodiment of the present invention aims to provide a method and device for detecting motion using a wireless LAN signal.

[0005] According to one embodiment of the present invention, a sensing device for detecting respiration of a measurement target using a measurement value indicating a channel state measured using a wireless LAN signal includes a memory; and a processor. The processor obtains a measurement value measuring the channel state based on a wireless LAN signal received by the sensing device from a sensing node, and detects respiration of a target within a sensing radius based on a cross-correlation operation between a predetermined base signal and a target signal in which the measurement value is listed in time series.

[0006] The pre-specified base signal may be a signal synthesized from multiple signals having different center frequencies.

[0007] The above processor can extract the frequency component that accounts for the largest proportion of the target component from the cross-correlation operation.

[0008] The above processor can detect the breathing of the measurement target based on the extracted frequency components.

[0009] The above processor can measure a channel state in each of a plurality of subcarriers included in the received wireless LAN signal.

[0010] The processor can generate a plurality of target signals based on a channel state measured in each of the plurality of subcarriers within one measurement window, and perform a cross-correlation operation on each of the plurality of signals and the base signal to determine whether the breathing of the target is detected.

[0011] According to one embodiment of the present invention, a method for operating a sensing device that performs respiration detection of a measurement target using a measurement value indicating a channel state measured using a wireless LAN signal includes the steps of: obtaining a measurement value that measures a channel state based on a wireless LAN signal received by the sensing device from a sensing node; and detecting respiration of a target within a sensing radius based on a cross-correlation operation between a predetermined base signal and a target signal in which the measurement values ​​are listed in time series.

[0012] The pre-specified base signal may be a signal synthesized from multiple signals having different center frequencies.

[0013] The step of obtaining a measurement value measuring a channel state based on a wireless LAN signal received by the sensing device from the sensing node may include a step of extracting a frequency component that accounts for the largest proportion of the target component through a cross-correlation operation.

[0014] The step of obtaining a measurement value obtained by measuring a channel state based on a wireless LAN signal received by the sensing device from the sensing node may further include a step of detecting the breathing of the measurement target based on the extracted frequency component.

[0015] The step of obtaining a measurement value obtained by measuring a channel state based on a wireless LAN signal received by the sensing device from the sensing node may include a step of measuring a channel state in each of a plurality of subcarriers included in the received wireless LAN signal.

[0016] The step of obtaining a measurement value obtained by measuring a channel state based on a wireless LAN signal received by the sensing device from the sensing node may further include a step of generating a plurality of target signals generated based on the channel state measured in each of the plurality of subcarriers within one measurement window, and a step of performing a cross-correlation operation on each of the plurality of signals and the base signal to determine whether the breathing of the target is detected.

[0017] One embodiment of the present invention provides a method and device for detecting motion using a wireless LAN signal.

[0018] FIG. 1 illustrates a system for detecting a user's motion using a wireless LAN signal according to an embodiment of the present invention.

[0019] FIG. 2 shows an AP, which is a sensing device, and a non-AP station, which is a sensing node, performing frame exchange according to an embodiment of the present invention.

[0020] FIG. 3 shows that channel status changes occur due to various events in a wireless channel in which a sensing device and a sensing node exchange signals according to an embodiment of the present invention.

[0021] FIG. 4 shows that a sensing device according to an embodiment of the present invention maintains a constant number of time series data through a plurality of measurement values ​​measured immediately before the most recent measurement value from a sensing node.

[0022] FIG. 5 shows an example of a procedure performed by a sensing device according to an embodiment of the present invention to detect presence.

[0023] FIG. 6 shows an example of extracting a value that is a characteristic of presence detection from a data set generated by a sensing device according to an embodiment of the present invention.

[0024] FIG. 7 shows a method for a sensing device according to an embodiment of the present invention to obtain features for each subcarrier.

[0025] FIG. 8 shows a method for determining presence from presence detection feature values ​​and threshold feature values ​​acquired by a device according to an embodiment of the present invention.

[0026] Figure 9 shows that the signal multipath changes and the received signal intensity changes due to the breathing of a person located within the sensing radius.

[0027] Figure 10 shows a simple model for wireless LAN signal propagation described in Figure 1.

[0028] FIG. 11 shows the magnitude of CSI of three subcarriers according to time collected in a situation where a person is positioned between a sensing device and a sensing node and repeats normal breathing and breathing pauses according to an embodiment of the present invention.

[0029] FIG. 12 illustrates a basic signal for cross-correlation operation according to an embodiment of the present invention.

[0030] Figure 13 shows the results of cross-correlation operations between the target signal and the element signal.

[0031] FIG. 14 shows that a sensing device according to an embodiment of the present invention maintains a constant number of time series data through a plurality of measurement values ​​measured immediately before the most recent measurement value.

[0032] Fig. 15 shows an operation of predicting a breathing rate from channel state information according to the explanation through Figs. 9 to 14.

[0033] FIG. 16 shows an operation of a sensing device according to an embodiment of the present invention to predict a breathing rate through data for one subcarrier composed of N time series.

[0034] The terms used in this specification have been selected from widely used and current terms, taking into account the functions of the present invention. However, these terms may vary depending on the intentions of those skilled in the art, customs, or the emergence of new technologies. Furthermore, in certain cases, the applicant may arbitrarily select terms, in which case their meanings will be described in the description of the relevant invention. Therefore, it should be noted that the terms used in this specification should be interpreted based on their substantive meaning and the overall content of this specification, rather than simply their names.

[0035] Throughout the specification, when a component is said to be "connected" to another component, this includes not only the case where the component is "directly connected," but also the case where the component is "electrically connected" with another component intervening therebetween. Furthermore, when a component is said to "include" a particular component, this does not exclude the other component, but rather allows the inclusion of other components, unless specifically stated otherwise. Furthermore, the terms "more than" or "less than" with respect to a specific threshold may be appropriately replaced with "greater than" or "less than", respectively, depending on the embodiment.

[0036]

[0037] FIG. 1 illustrates a system for detecting a user's motion using a wireless LAN signal according to an embodiment of the present invention.

[0038] An AP can be associated with various non-AP stations. The AP can detect movement around the AP and the station based on the measurement results of RF signals transmitted by the non-AP stations or the measurement results of signals transmitted by the stations. Specifically, when movement occurs around the AP and the station, the signal measurement results change. Therefore, a sensing node or AP can detect movement around the AP and the station by tracking the change in the signal measurement results. At this time, the signal measurement results may be channel state information (CSI). Additionally, the CSI can be obtained through the LTF (Tong Training Field) of the WLAN PPDU. In this specification, a station that transmits a signal used for signal measurement is referred to as a sensing node, and a device that performs signal measurement is referred to as a sensing device.

[0039] The sensing device can calculate the detected motion level based on the received signal measurement results. At this time, the motion level is referred to as an activity level. The sensing node or sensing device transmits the calculated activity level to a cloud server. At this time, the sensing node or sensing device can also transmit wireless LAN information related to the signal measurement. The wireless LAN information may include at least one of information about the RF signal used for signal measurement, information about the basic service set (BSS) to which the sensing node or sensing device belongs, and information about the station connected to the BSS.

[0040] The user application can display information received from the cloud server based on user input. Specifically, the user application can display activity levels calculated from signal measurement results received from sensing nodes. The user application can be a mobile application running on a smartphone or a web application.

[0041] The cloud server stores data received from a sensing device or sensing node. Furthermore, the cloud server processes user application requests and transmits responses to the user application requests. Furthermore, the cloud server can transmit commands to the sensing node or sensing device to which the user application has sent the request. Users can check information about stations connected to the AP through the user application. The station information may include at least one of the station's name and location.

[0042] In FIG. 1, the device for determining motion detection based on measurement results is described as an AP. However, in another specific embodiment, a cloud server may determine motion detection based on measurement results. In another specific embodiment, a sensing node may determine motion detection based on measurement results. In this case, motion detection may include at least one of occupancy detection, which detects whether a user is located in a specific space, or respiration detection, which detects the user's breathing.

[0043]

[0044] FIG. 2 shows an AP, which is a sensing device, and a non-AP station, which is a sensing node, performing frame exchange according to an embodiment of the present invention.

[0045] Channel state information (CSI) can indicate signal distortion as it propagates from a transmitter to a receiver. CSI can also reflect effects such as scattering, fading, and power attenuation over distance. Transmitters and receivers can respond to changes in channel conditions based on CSI. This allows them to maintain high reliability in high-data-rate, multi-antenna communications. Because CSI can indicate signal distortion as it propagates from a transmitter to a receiver, it must be measured at the receiver. The measurement process must be performed continuously and at appropriate times. Therefore, it is recommended that the measurement be performed by a device powered by a non-battery source.

[0046] In Fig. 2, the AP transmits a null data packet to a sensing node, and a non-AP station (STA), which is a sensing node, transmits an ACK frame to the AP. The AP measures CSI for a signal including the ACK frame. The AP can transmit the QoS null frame using the maximum bandwidth supported by the AP or the non-AP station. The non-AP station can transmit the ACK frame to the AP using a bandwidth equal to the bandwidth of the received QoS null frame. However, the non-AP station may transmit the ACK frame using a bandwidth narrower than the maximum bandwidth due to environmental factors, such as the channel environment and the strength of the received signal.

[0047] This allows the AP to minimize the impact on the battery power consumption of non-AP stations. Furthermore, the CSI measured by the AP can indicate the state of the wireless channel at the time of ACK frame transmission. This state of the wireless channel indicates signal propagation distortion during the wireless transmission of the ACK frame transmitted by the sensing node to the AP. As previously described, the non-AP station can be the sensing device, and the AP can be the sensing node. In this case, the non-AP station can perform the operations of the AP in FIG. 2, and the AP can perform the operations of the non-AP station.

[0048]

[0049] FIG. 3 shows that channel status changes occur due to various events in a wireless channel in which a sensing device and a sensing node exchange signals according to an embodiment of the present invention.

[0050] When there is no obstruction between the sensing device and the sensing node, the two can exchange signals through the optimal path. The optimal path between the sensing device and the sensing node can be disrupted by human movement. This can cause both the sensing device and the sensing node to adjust the signal path.

[0051] The length of the propagation path may be the shortest along the line of sight. It may be proportional to the distance of the obstacle reflecting the signal. As shown in (b) of Fig. 3, based on the signal received through the line of sight path, the signal received through reflection has a time delay in the time domain. The signal y1 that reaches the receiver through the l1 path of Fig. 3 (a) is l los than the path length Long, and accordingly A phase difference of the order of l can occur. The signal y2 reaching the receiver through the l2 path is l los than the path length Long, and accordingly A phase difference of that magnitude may occur. The received signal may be the sum of the signals transmitted through all possible paths. In addition, wireless signals are attenuated depending on the length of the propagation path and the type of reflective material. The longer the propagation path, the greater the degree of signal attenuation. In an ideal static space, a wireless signal received does not change over time. On the other hand, a signal that is wirelessly propagated and received in a space where an object moves may change according to the object's movement. Therefore, the movement of an object may cause a change in the received signal. In the present invention, the range of the object's position where the object's movement sufficiently changes the change in the received signal is referred to as the sensing radius. Presence detection also starts from detecting signal changes due to the movement of a person in a space. However, for occupancy detection, the sensing device must detect subtle signal changes over a long period of time, unlike general motion detection. This will be explained with reference to FIGS. 4 to 8.

[0052]

[0053] FIG. 4 shows that a sensing device according to an embodiment of the present invention maintains a constant number of time series data through a plurality of measurement values ​​measured immediately before the most recent measurement value from a sensing node.

[0054] A sensing device can use a value representing changes such as reflection, absorption, and attenuation that a signal experiences as it propagates from a transmitter to a receiver, such as CSI, as a measurement value. The sensing device can obtain the sum of N-1 measurement values ​​measured immediately before the most recent value and the most recent measurement value. The sensing device can generate a data set used as an input for an algorithm based on the obtained values. In addition, the sensing device can obtain measurement values ​​by measuring the channel status at regular time intervals. For convenience of explanation, the time period for obtaining N measurement values ​​is referred to as a measurement window. In addition, N may be a predetermined value or may be adjusted by the sensing device. For example, N may be a square number greater than 9. In addition, N may change in proportion to the window and the measurement interval. Accordingly, if the length of the data set measured immediately before the most recent value is N, the sensing device can generate a data set by deleting the oldest measured data among the measurement values ​​constituting the data set and adding the most recently measured value. In a specific embodiment, if the window is 10 seconds and the measurement interval is 0.1 second, N can be 100. As described in FIG. 2, to maintain consistency of values ​​measured with different bandwidths, the data within the window can be composed of CSI values ​​of the primary channel. The measurement values ​​measured by the sensing device increase in proportion to the number of receiving antennas of the sensing device. Therefore, the number of data sets can be a value proportional to the number of receiving antennas.

[0055]

[0056] FIG. 5 shows an example of a procedure performed by a sensing device according to an embodiment of the present invention to detect presence.

[0057] The measured data set described through Fig. 4 can be divided into a feature data set for occupancy detection and a data set for threshold through the input data set generation block. The lengths of both data sets can be squares of integers. The length of the occupancy detection feature data set can be greater than the length of the threshold feature data set. Specifically, the length of the occupancy detection feature data set can be 100, and the length of the threshold feature data set can be 25. The logical operations performed in the occupancy detection feature calculation block and the logical operations of the threshold feature calculation block can be the same. The occupancy detection feature block can calculate occupancy detection features, and the threshold feature calculation block can calculate threshold features. The occupancy detection features represent the features of the data set for determining occupancy and absence. The sensing device can use the threshold features as threshold values ​​that serve as criteria for determining occupancy and absence. An embodiment of the feature calculation block is described through Fig. 6. The occupancy detection features and the threshold features are input to the occupancy determination block. The presence determination block uses presence detection features and threshold features to ultimately determine presence or absence within the sensing radius. An embodiment of the presence determination block is described with reference to FIG. 8.

[0058]

[0059] FIG. 6 shows an example of extracting a value that is a characteristic of presence detection from a data set generated by a sensing device according to an embodiment of the present invention.

[0060] The sensing device can generate a CSI data set composed of a two-dimensional matrix according to a preset measurement window. At this time, M is the number of CSIs corresponding to subcarriers, and N can be the number of samples corresponding to a preset window size. N can be determined according to the operation of the presence detection feature value and the operation of the threshold feature detection value. The two-dimensional matrix data can be classified into data in the form of an array for each subcarrier. The array data includes the CSI value for each subcarrier as a time series. The sensing device composes the array data into N CSIs corresponding to the window size, and the size x It can be reconstructed into a matrix. At this time, the matrix is ​​arranged in the order of the time the data was collected, from (1, 1) to ( , ) is composed of (1, 1) and ( , ) may be the most recently measured value. The sensing device can obtain a value that is a feature of the presence detection from the reconstructed matrix. Specifically, the sensing device can calculate a value that is a feature of the presence detection from the data set using the Get_SC_feature block of FIG. 6. An embodiment of the operation of the Get_SC_feature block is described with reference to FIG. 7. The sensing device performs a feature operation of the presence detection for all subcarriers that will use the measured values. At this time, the sensing device can perform the operation for one subcarrier and then perform the operation for the next subcarrier. In another specific embodiment, the sensing device can perform the feature operation of the presence detection for all subcarriers that will use the measured values ​​in parallel. The sensing device can obtain the final feature value (CS_feature) by calculating the cumulative product for all calculated feature values ​​(CS_feature) for each subcarrier and taking the absolute value. When the sensing device has two or more receiving antennas, the sensing device can obtain the final feature value by averaging the feature values ​​calculated for each antenna. Feature values ​​may be unitless. Furthermore, the range of feature values ​​can be adjusted through linear transformation. Specifically, the sensing device can adjust the range of feature values ​​using common logarithms.

[0061]

[0062] FIG. 7 shows a method for a sensing device according to an embodiment of the present invention to obtain features for each subcarrier.

[0063] In the description through Fig. 6, the time series data reconstructed into a two-dimensional matrix may have values ​​that are temporally adjacent in the row direction and values ​​with relatively longer time intervals in the column direction. The sensing device may obtain the final feature value by using the standard deviation of each measurement window having a first interval and the standard deviation of a plurality of measurement windows having a second interval in one subcarrier. At this time, the first interval may be smaller than the second interval. Specifically, the second interval may be a multiple of the first interval. Specifically, the sensing device may calculate the standard deviation for each row and column of the data set. In a specific embodiment, the sensing device may obtain a standard deviation array for each row and a standard deviation array for each column in the two-dimensional matrix. The standard deviation value for the first row may be expressed as R1, and the standard deviation value for the first column may be expressed as C1. Ideally, for data collected in a state where there is no movement within the sensing area, C n Wow R n can have the same value. C if data is collected while movement occurs within the sensing area. n Wow R n can have different values. The sensing device calculates C from each row and column to obtain the feature value (CS_feature) for each subcarrier. n Wow R n About can be performed. At this time, represents the absolute value of x. Therefore, the feature value for each subcarrier can have a very small value in a static environment and a relatively large value even in an environment where slight movement occurs.

[0064]

[0065] FIG. 8 shows a method for determining presence from presence detection feature values ​​and threshold feature values ​​acquired by a device according to an embodiment of the present invention.

[0066] The sensing device can obtain a pure feature value (pure_feature) composed of positive and negative numbers based on 0 by subtracting a threshold feature value from the occupancy detection feature value. At this time, a large value in the positive direction may mean a high probability of occupancy, and a large value in the negative direction may mean a high probability of absence. The pure feature value may be scaled through integer multiplication. In a specific embodiment, the integer may be 5. The scaled value may have a limited range. Specifically, a value less than or equal to -100 may be converted into a scaled feature value (scaled_feature) that is scaled to -100, and a value greater than or equal to 100 may be converted into a scaled feature value. The sensing device may apply a moving average to the scaled feature value.

[0067] The sensing device can extract a smoothed feature value (smoothed_feature) by applying a moving average with a window size of a predetermined size, for example, 100. The sensing device can obtain a presence probability value (presence_prob) using the adjusted feature value and the smoothed feature value. If the adjusted feature value is 100, the sensing device can determine the presence probability value as 100, and if the adjusted feature value is less than or equal to 0, the presence probability value can be determined as 0. If neither condition is satisfied, the sensing device can set the presence probability value to the smoothed feature value. Accordingly, the presence probability value can have a value between 0 and 100. The sensing device can determine the final presence by comparing the calculated presence probability value with a threshold value x. X can be a predetermined value or can be adjusted by the sensing device or the user. For example, if x is set to 0, and the presence probability value is positive, it can be determined as presence, and if it is 0, it can be determined as absence.

[0068] A sensing device according to an embodiment of the present invention can obtain an occupancy detection feature value and a threshold feature value, and determine whether a detection target is present based on the occupancy feature value and the threshold feature value. At this time, the sensing device can obtain the occupancy detection feature value in the corresponding subcarrier using the standard deviation of each measurement window having a first interval in one subcarrier and the standard deviation of multiple measurement windows having a second interval. At this time, the first interval may be smaller than the second interval. Specifically, the second interval may be a multiple of the first interval. In addition, the sensing device can compare the occupancy detection feature value and the threshold feature value in the multiple subcarriers, and determine whether a target is present based on the comparison result. The sensing device can obtain a measurement value for each of the multiple measurement windows in each of the multiple subcarriers. Specifically, the sensing device can obtain the occupancy detection feature value through the embodiments described with reference to FIGS. 5 to 7. In addition, the sensing device can obtain the threshold feature value through the embodiments described with reference to FIG. 8. In addition, in these embodiments, the measurement value may be a CSI value of a primary channel. At this time, the multiple carriers may be multiple return carriers included in the main channel.

[0069]

[0070] Figure 9 shows that the signal multipath changes and the received signal intensity changes due to the breathing of a person located within the sensing radius.

[0071] When a person inhales while breathing, the chest or abdomen expands, and when they exhale, the chest or abdomen contracts. Figure 9 (a) shows the change in the multipath along which a signal is propagated due to the expansion and contraction of the chest or abdomen. For example, the multipath may change from l1 to l2, or from l2 to l1 due to the expansion and contraction of the chest or abdomen. Based on the contents described through Figures 1 and 2, when signals are continuously exchanged between a sensing node and an AP during breathing, the magnitude of the signal received by the AP can be expressed as in Figure 9 (b). The value of the y-axis represents the magnitude of the received signal for a unit signal. At this time, the periodicity of the pattern change in the magnitude of the continuously received signal may be the same as the periodicity of the breathing pattern.

[0072]

[0073] Figure 10 shows a simple model for wireless LAN signal propagation described in Figure 1.

[0074] The received signal y is the sum of the signals that the transmitted signal x passes through multiple paths and reaches the receiving antenna. H can represent a value that represents the changes in reflection, absorption, and attenuation that the signal experiences as it propagates from the transmitter to the receiver. The transmitter transmits a mutually known signal x that has been agreed upon with the receiver. The receiver can infer the channel state by dividing the signal x from the received signal y. In wireless communications, this is called Channel State Information (CSI). In an environment where the channel state fluctuates greatly, CSI can be used to estimate the wireless channel state and restore distorted signals. Since CSI represents the distortion that a signal has as it propagates, the breathing characteristics described in FIG. 10 can be reflected. Since WLAN signals use the Orthogonal Frequency Division Multiplexing (OFDM) modulation technique, CSI can include values ​​equal to the maximum number of subcarriers for one packet. The CSI value corresponding to each subcarrier can be a complex number. In addition, the value of CSI can vary depending on the number of CSI transmitting and receiving antennas, as explained above.

[0075]

[0076] FIG. 11 shows the magnitude of CSI of three subcarriers according to time collected in a situation where a person is positioned between a sensing device and a sensing node and repeats normal breathing and breathing pauses according to an embodiment of the present invention.

[0077] The sensing device transmits a QoS null frame to the sensing node every 100 ms, and the sensing node transmits an Ack frame to the AP in response to this, showing the acquired CSI. Figure 11 shows the CSI for the first, middle (DC+1), and last subcarriers among all collected CSI. As explained in Figure 9, the expansion and contraction of the chest and abdomen during breathing causes changes in the multipath through which the signal propagates. The change in the multipath causes the CSI size to change. Depending on the embodiment, the CSI size may decrease or increase when exhaling and may increase or decrease when inhaling. On the other hand, the CSI size remains relatively constant during periods without breathing. In wireless LAN communication, OFDM uses 20 MHz as the basic bandwidth, and 20 MHz can be composed of 64 subcarriers with a bandwidth of 312.5 kHz. Therefore, each subcarrier has a different center frequency and a different wavelength. Accordingly, even in the same received CSI, different CSI size changes may occur due to constructive or destructive interference for each subcarrier.

[0078]

[0079] A sensing device can detect the respiration of a subject within a sensing radius based on the cross-correlation operation value between a pre-specified base signal and a target signal, which is a time-series listing of measured measurement values. Through this, the sensing device can extract the frequency component that accounts for the largest proportion of the target signal through the cross-correlation operation of the base signal and the target signal. The sensing device can detect respiration based on the largest frequency component. Through this, the sensing device can determine the frequency of the extracted frequency component as a heart rate. In this case, the pre-specified base signal may be the sum of multiple signals having different center frequencies. In addition, the measurement value may be the CSI described above.

[0080] The cross-correlation operation of the detailed sensing device is explained through Figs. 12 to 16.

[0081] FIG. 12 illustrates a basic signal for cross-correlation operation according to an embodiment of the present invention.

[0082] The base signal is composed of multiple frequency signals. In the embodiment of Fig. 12, the frequencies of the signals are illustrated as component signals connected in series from 1 rpm (revolutions per minute) to 60 rpm. In Fig. 12, the x-axis represents samples, and the y-axis represents magnitude. The component signals of different frequencies that constitute the base signal may include signals of at least one cycle. Since the base signal is for cross-correlation with the time series CSI, the minimum length of each frequency signal is the number of target CSI samples. As described above, the CSI composed of the time series to be cross-correlated with the base signal is referred to as the target signal. For example, a target signal with a breathing rate of 12 bpm (breathing per minute), 10 CSIs acquired per second, and a length of 10 seconds may be a signal in which a waveform composed of 100 samples is repeated twice. The phase of the target signal may vary depending on the CSI acquisition time and the breathing time. Therefore, the 12 rpm element signal constituting the basic signal is repeated at least 3 times so that the target signal of any phase can be included in the element signal. For example, if the breathing rate is 18 bpm, the target signal corresponding to 10 seconds has a waveform repeated 3 times. Similarly, so that the target signal of any phase can be included in the element signal, the 18 rpm element signal is repeated at least 4 times. In this way, the minimum repetition period of the element signal can be set so that the target signal of any phase can be included. In addition, the minimum repetition period of the element signal can vary depending on the CSI acquisition cycle and period constituting the target signal. Fig. 11 is a basic signal in the form of a monotonically increasing frequency according to an embodiment, and the element signals constituting each frequency do not need to be combined in order.

[0083]

[0084] Figure 13 shows the results of cross-correlation operations between the target signal and the element signal.

[0085] Figure 13(a) shows the cross-correlation function of the target signal and the base signal collected in an environment with normal breathing. Figure 13(b) shows the cross-correlation function of the target signal and the base signal collected in an empty space environment. As shown in Figure 13(a), the cross-correlation function of the target signal and the base signal in an environment with breathing has a distinct maximum value. On the other hand, as shown in Figure 13(b), the maximum value of the cross-correlation function of the target signal and the base signal in an empty space environment is not distinct. The larger the value of the cross-correlation function, the higher the correlation between the two signals. Therefore, prediction can be made based on the frequency that maximizes the cross-correlation function. At this time, the frequency at which the value of the cross-correlation function is maximum may correspond to the breathing rate indicated by the target signal.

[0086]

[0087] FIG. 14 shows that a sensing device according to an embodiment of the present invention maintains a constant number of time series data through a plurality of measurement values ​​measured immediately before the most recent measurement value.

[0088] The sensing device can use the value obtained based on the sum of the N-1 measurement values ​​measured immediately before the most recent value and the most recent measurement value as the feature value of the respiration detection algorithm. At this time, N is a natural number greater than 1. In addition, the sensing device can obtain the measurement value by measuring the channel status at regular time intervals. For convenience of explanation, the time period for obtaining N measurement values ​​is referred to as a measurement window. In addition, N can be a pre-specified value or can be adjusted by the sensing device. Therefore, N can change depending on the size of the measurement window and the measurement interval. For example, if the measurement window is 10 seconds and the measurement interval is 0.1 second, N can be 100. The data within the measurement window can be composed of CSI measured from the primary channel.

[0089]

[0090] Fig. 15 shows an operation of predicting a breathing rate from channel state information according to the explanation through Figs. 9 to 14.

[0091] The input CSI may be N data composed of CSI measured on the main channel as described through Fig. 14. In Fig. 14, #sc represents the number of subcarriers constituting the CSI. The sensing device performs breathing rate prediction for each data composed of N time series for each subcarrier. Breathing rate prediction (Predict breathing rate per subcarrier) using data for each subcarrier is described through Fig. 16. The sensing device temporarily stores the breathing rate prediction value (pbr_sc) for each subcarrier in an array. Thereafter, the sensing device stores non-zero values ​​among the values ​​in the array in an array (cand_br; candidate of breathing rate). If the length of cand_br is less than a preset T, the temporary breathing rate (temp_rate) is set to 0. Otherwise, the sensing device sets the intermediate value of cand_br as temp_rate. Therefore, the sensing device can select the value of temp_rate as a representative breathing rate prediction value for the N time series data. Here, T is a natural number greater than 0. Since normal breathing maintains a constant breathing rate, the breathing rate prediction value calculated from the continuous time series can be a non-zero value continuously over time. Therefore, the sensing device can select the final sensing result based on the results sensed for a certain period of time. If the maximum number of consecutive non-zero elements in the consecutive M representative breathing rate prediction values ​​BR is greater than αM, the sensing device can determine the final breathing rate prediction value as BR[1]. Here, α is a real number between 0 and 1. If the maximum number of consecutive non-zero elements belonging to the consecutive M representative breathing rate prediction values ​​BR is less than αM, the sensing device can determine the final breathing rate prediction value as 0. For example, the maximum number of consecutive non-zero elements belonging to the array [1, 3, 5, 0, 7, 8] is 3.

[0092]

[0093] FIG. 16 shows an operation of a sensing device according to an embodiment of the present invention to predict a breathing rate through data for one subcarrier composed of N time series.

[0094] In Fig. 16, sc_CSI is data for one subcarrier, and base_signal is the base signal. base_idx is an array containing frequency information of the base signal. The length of base_idx is equal to the length of the base signal, and the array element values ​​are the frequencies of the base signal. For example, base_idx for a base signal consisting of low to high frequencies may have a step-like form. The size of the CSI may vary depending on the collection environment. Therefore, a normalization process is necessary to overcome the dependence on the measurement environment. The sensing device can z-score normalize the sc_CSI. The z-score normalized data may have a value between +2 and -2. The normalized data may be the target signal described in Fig. 13. The sensing device can obtain a cross-correlation function by cross-correlating the target signal and the base signal. Thereafter, the sensing device can obtain the maximum value of the cross-correlation function and the input value that maximizes it. If the maximum value exceeds a preset threshold, the sensing device can determine that respiration has been detected. At this time, the sensing device can obtain the predicted respiration rate (pbr_sc) by inputting the input value that maximizes the value of the cross-correlation function into base_idx. If the maximum value is less than or equal to a preset threshold, the sensing device determines that respiration is not detected and sets pbr_sc to 0. The preset threshold may be determined based on at least one of a normalization technique or the length of the target signal.

[0095] In the previously described examples of respiration detection, the cross-correlation function was used because it can roughly estimate beats per minute (bpm). Furthermore, compared to other algorithms such as the fast Fourier transform (FFT), the cross-correlation function can effectively extract the frequency components that account for the largest proportion even in relatively short intervals. For example, frequency transformation using the FFT requires at least 60 seconds of target signal measurement to achieve a resolution of 1 rpm (0.01667 Hz).

[0096]

[0097] While the present invention has been described using wireless LAN communication as an example, it is not limited thereto and can be equally applied to other communication systems, such as cellular communication. Furthermore, while the methods, devices, and systems of the present invention have been described with reference to specific embodiments, some or all of the components and operations of the present invention can be implemented using a computer system with a general-purpose hardware architecture.

[0098] The features, structures, effects, etc. described in the embodiments above are included in at least one embodiment of the present invention, and are not necessarily limited to just one embodiment. Furthermore, the features, structures, effects, etc. exemplified in each embodiment can be combined or modified in other embodiments by those skilled in the art to which the embodiments pertain. Therefore, the contents related to such combinations and modifications should be construed as being included within the scope of the present invention.

[0099] Although the above has been described focusing on embodiments, these are merely examples and do not limit the present invention. Those skilled in the art to which the present invention pertains will appreciate that various modifications and applications not exemplified above are possible without departing from the essential characteristics of the present embodiment. For example, each component specifically shown in the embodiments can be modified and implemented. In addition, differences related to such modifications and applications should be interpreted as being included within the scope of the present invention defined in the appended claims.

Claims

1. In a sensing device that detects the respiration of a measurement target using a measurement value indicating a channel status measured using a wireless LAN signal, memory; and Contains a processor, The above processor Obtain a measurement value measuring the channel status based on a wireless LAN signal received from the sensing node to the sensing device, Detecting the breathing of a target within a sensing radius based on a cross-correlation operation between a pre-specified base signal and a target signal that lists the above measurement values ​​in time series. Sensing device.

2. In paragraph 1, A pre-specified base signal is a signal that is a composite of multiple signals with different center frequencies. Sensing device.

3. In paragraph 1, The above processor Extracting the frequency component that accounts for the largest proportion of the target component from the cross-correlation operation Sensing device.

4. In paragraph 3, The above processor Detecting the breathing of the measurement subject based on the frequency components extracted above Sensing device.

5. In paragraph 1, The above processor Measuring the channel state in each of the multiple subcarriers included in the received wireless LAN signal Sensing device.

6. In paragraph 5, The above processor Generating a plurality of target signals based on the channel state measured in each of the plurality of subcarriers within one measurement window, The cross-correlation operation is performed on each of the plurality of signals and the base signal to determine whether the breathing of the subject is detected. Sensing device.

7. In a method of operating a sensing device that detects respiration of a measurement target using a measurement value indicating a channel status measured using a wireless LAN signal, A step of obtaining a measurement value obtained by measuring a channel state based on a wireless LAN signal received by the sensing device from the sensing node; and A step of detecting the breathing of a target within a sensing radius based on a cross-correlation operation between a pre-specified base signal and a target signal that lists the measurement values ​​in time series. How it works.

8. In paragraph 7, A pre-specified base signal is a signal that is a composite of multiple signals with different center frequencies. How it works.

9. In paragraph 7, The step of obtaining a measurement value obtained by measuring a channel state based on a wireless LAN signal received by the sensing device from the sensing node is A step of extracting the frequency component that accounts for the largest proportion of the target component from the cross-correlation operation is included. How it works.

10. In paragraph 9, The step of obtaining a measurement value obtained by measuring a channel state based on a wireless LAN signal received by the sensing device from the sensing node is Further comprising a step of detecting the breathing of the measurement target based on the extracted frequency components. How it works.

11. In paragraph 7, The step of obtaining a measurement value obtained by measuring a channel state based on a wireless LAN signal received by the sensing device from the sensing node is A step of measuring a channel state in each of a plurality of subcarriers included in the received wireless LAN signal is included. How it works.

12. In paragraph 11, The step of obtaining a measurement value obtained by measuring a channel state based on a wireless LAN signal received by the sensing device from the sensing node is A step of generating a plurality of target signals based on the channel state measured in each of the plurality of subcarriers within one measurement window; and Further comprising a step of determining whether the breathing of the subject is detected by performing a cross-correlation operation on each of the plurality of signals and the base signal. How it works.

Images