Human Presence Sensing Method and System Based on In-Vehicle WiFi

Abstract

The method and system for human presence perception based on in-vehicle WiFi proposed by the present invention include: a receiving device receives a radio frequency signal emitted from a transmitting device; each receiving antenna of the receiving device receives the channel state information CSI of multiple subcarriers, and calculates the energy spectrum of the CSI signal on each subcarrier; according to the energy spectrum of the CSI signal within a period window, calculates the variance of the CSI energy spectrum of each subcarrier; determines whether the mean value of the CSI energy spectrum variances of all obtained subcarriers is higher than a preset body movement threshold. When the determination result is yes, it is determined that there is someone in the car; when the determination result is no, the energy spectrum signal of the CSI of each subcarrier is processed according to a preset method, and it is determined whether there is someone in the car by judging whether the peak point of the breathing signal spectrum is within the human breathing frequency range. The present invention can realize the perception of human presence in the car based on a simple single-antenna WiFi transceiver, and accurately detect the human state.

Description

Technical Field

[0001] The present invention relates to the field of in-vehicle wireless sensing technology, and more particularly, to a method and system for human presence perception based on in-vehicle WiFi.

Background Art

[0002] In the enclosed space of a vehicle, the temperature can reach as high as 60°C in a short time in summer. If a child is accidentally left alone in the vehicle and not discovered in time, dangerous events such as death may occur. To avoid such tragedies, it is very necessary to use human presence perception technology to detect the rear-seat passengers in the vehicle and prevent adverse consequences through timely alarm. With the rapid development of the Internet of Things technology, in-vehicle WiFi has been gradually popularized, and non-contact activity perception based on wireless signals has received extensive attention from people, and there have been many related studies and applications. Non-contact activity perception refers to the ability to detect whether there is a moving object in the environment without the need for the target to be sensed to carry any device.

[0003] For non-contact activity perception, researchers have proposed using the correlation of the amplitude of the Channel State Information (CSI) in the WiFi device in the time domain to sense and detect whether there is a moving object. When there is someone moving in the environment, there are obvious changes in the amplitude of the channel state information, and the correlation in the time domain is relatively low. Using this feature, the difference in the amplitude correlation of the channel state information when there is human movement in the environment can be learned in advance, so as to determine whether someone is moving according to the correlation of the CSI signal amplitude obtained in the actual test. However, this method requires pre-learning the environment, and when the environment changes, the device parameters need to be re-calibrated, and it is impossible to sense and detect a stationary human body. A method and system for indoor activity detection based on wireless signals disclosed in Chinese Patent CN109257126A propose to process the CSI data of two antennas at the WiFi signal receiving end to obtain the Doppler frequency shift spectrum, and determine whether there is a moving object according to whether the energy of the highest spectral peak of the Doppler frequency shift spectrum is higher than a preset threshold; at the same time, for different environments, pre-learning and calibration are not required. However, similar to the previous method, this method also cannot sense and detect a stationary human body (such as sleeping), and it is impossible to detect the presence of a human body when someone is not moving or in a quasi-static (sleeping) state, and it can only detect a moving human body, and it is not applicable to the scenario where there is a child sleeping in the back seat. And it requires two receiving antennas, with high device requirements and not applicable to the existing in-vehicle Internet of Things scenario.

[0004] It can be seen that due to the frequent changes in the initial environment of the enclosed space inside the vehicle (such as storing sundries in the vehicle irregularly, etc.) and factors such as vehicle-mounted equipment limitations, the existing technologies are not applicable to the enclosed space inside the vehicle. Therefore, how to solve the problem of sensing human presence in the enclosed space inside the vehicle is a technical issue worthy of great attention.

Summary of the Invention

[0005] In order to overcome the deficiencies of the above-mentioned existing technologies, the present invention provides a method and system for sensing human presence based on in-vehicle WiFi, which can realize the sensing of human presence inside the vehicle based on a simple single-antenna WiFi transceiver, and accurately detect the presence or absence of a human being in both the moving state and the quasi-static state.

[0006] On the one hand, the present invention proposes a method for sensing human presence based on in-vehicle WiFi, including:

[0007] Step 1: A receiving device receives a radio frequency signal emitted from a transmitting device, and the receiving device has one or more receiving antennas;

[0008] Step 2: Obtain the channel state information CSI of multiple subcarriers according to the radio frequency signals received by each receiving antenna, and calculate the energy spectrum of the CSI signal on each subcarrier;

[0009] Step 3: Calculate the variance of the CSI energy spectrum of each subcarrier according to the energy spectrum of the CSI signal within a time window;

[0010] Step 4: Determine whether the mean value of the variances of the CSI energy spectra of all obtained subcarriers is higher than a preset body movement threshold. When the judgment result is yes, it is determined that there is someone in the car, and go to Step 6; when the judgment result is no, process the energy spectrum signal of the CSI of each subcarrier according to a preset method to obtain a respiration signal spectrum;

[0011] Step 5: Determine whether the peak point of the obtained respiration signal spectrum is within the human respiration frequency range. When the judgment result is yes, it is determined that there is someone in the car; when the judgment result is no, it is determined that there is no one in the car;

[0012] Step 6: End.

[0013] Further, based on the above technical solution, the radio frequency signal is a WiFi data packet emitted from the same transmitting device, and the transmitting device includes an in-vehicle WiFi device.

[0014] Further, on the basis of the above technical solution, the calculation of the energy spectrum of the CSI signal on each subcarrier includes:

[0015] Set the system sampling rate to F s, with the unit of Hz, the channel state information on the sub - carrier f at time t is calculated as H(t, f), and the energy spectrum G(t, f) is calculated according to the following formula:

[0016] G(t,f) = |H(t, f)| 2 , where H(t, f) is a complex number.

[0017] Furthermore, based on the above - mentioned technical solution, calculating the variance of the CSI energy spectrum of each sub - carrier includes:

[0018] Set a time window size of W1, with the unit of seconds. The channel state information within the time window is all the sampling points within this time interval, that is, the sampling points between seconds and t seconds; since the system sampling rate is F s , so there are a total of W1 * F s sampling points within the time window. For each sub - carrier, calculate the variance of the CSI energy spectrum Var(f) within its corresponding time window according to the following formula:

[0019] where is the mean value of the CSI energy spectrum on sub - carrier f within the time window.

[0020] Furthermore, based on the above - mentioned technical solution, when comparing the mean value of the CSI energy spectrum variances of all sub - carriers with a preset body movement threshold d1, if the judgment result is no, it further includes: updating the preset body movement threshold d1.

[0021] Furthermore, based on the above - mentioned technical solution, updating the preset body movement threshold d1 includes:

[0022] Set a time window W2, with the unit of seconds, and judge whether the standard deviation of the mean value of the energy spectrum variances of all sub - carriers within the time window W2 is higher than the update threshold d2. When the judgment result is yes, do not update the body movement threshold d1; when the judgment result is no, update it to a new body movement threshold d1.

[0023] Furthermore, based on the above - mentioned technical solution, the new body movement threshold d1 is the sum of the mean value of the CSI energy spectrum variances of all sub - carriers within the time window W2 plus three times its standard deviation.

[0024] Furthermore, based on the above - mentioned technical solution, processing the energy spectrum signal of each sub - carrier CSI according to a preset method to obtain a respiratory signal spectrum includes: converting the time - domain CSI energy spectrum signal within a time window W3 into a corresponding frequency - domain signal through fast Fourier transform FFT, and finding the frequency point corresponding to its maximum amplitude. The unit of the time window W3 is seconds.

[0025] Further, based on the above technical solution, determining whether the peak point of the respiration signal spectrum is within the human respiration frequency range includes: calculating the corresponding frequency according to the frequency point corresponding to the found maximum amplitude; if the frequency is within the interval of the normal human respiration frequency range, it is determined that there is someone in the vehicle, where the normal human respiration frequency range is 0.15 Hz - 0.50 Hz.

[0026] Further, based on the above technical solution, the sampling rate of the system is 10 Hz.

[0027] Further, based on the above technical solution, W1 is 2.

[0028] Further, based on the above technical solution, W2 is 60.

[0029] Further, based on the above technical solution, W3 is 10.

[0030] On the other hand, the present invention also proposes a human presence perception system based on in-vehicle WiFi, including:

[0031] A transmitting device, a receiving device, a processor, a memory, and a controller. The controller controls the transmitting device and the receiving device to transmit and receive radio frequency signals. The memory stores a medium with program code. When the processor reads the program code stored in the medium, the system can execute the method according to any one of claims 1 - 13.

[0032] Based on the inventive concept of the present invention, the following beneficial technical effects can be obtained:

[0033] The technical solution disclosed by the present invention can realize the perception of human presence in a car only by using a single - antenna WiFi transceiver device. By obtaining the channel state information of all sub - carriers of the receiving antenna and calculating the relevant statistical characteristics of its energy spectrum, the perception of the human motion state is realized; and through the online update of the body movement threshold, it is easier to be applicable to different deployment environments and hardware devices.

[0034] In addition, the present invention also adopts a perception technology based on respiration frequency, which can realize the perception of the human body in a quasi - static state, is more suitable for the scenario where a child in the back row of a car falls asleep and is forgotten, and at the same time, the hardware requirements of the device are not high, the deployment is convenient, the operability is strong, and the application range is wider.

Description of the Drawings

[0035] Figure 1 It is a flowchart of a human presence perception method based on in - vehicle WiFi according to an embodiment of the present invention.

Detailed Embodiments

[0036] For ease of understanding, this specific embodiment is a preferred embodiment of the human presence sensing method and system based on in-vehicle WiFi proposed by the present invention, to illustrate in detail the structure and inventive points of the present invention, but does not serve as the limited protection scope of the claims of the present invention.

[0037] As a preferred embodiment, in an in-vehicle environment, it includes a transmitting device and a receiving device. The transmitting device is used to send radio frequency signals, and the receiving device is used to receive radio frequency signals. The controller in the in-vehicle computer can control the transceiver of the radio frequency signals of the transmitting device and the receiving device. The processor in the in-vehicle computer can control and sense whether there is a human body in the vehicle by reading the program code stored in the storage medium of the in-vehicle computer for the radio frequency signals received and sent.

[0038] See Figure 1 The operation flowchart of the preferred embodiment of the present invention, including:

[0039] S1: Send radio frequency signals to the receiving device through in-vehicle WiFi and collect the corresponding channel state information CSI. Specifically, the in-vehicle WiFi transmitting device sends WiFi data packets at a speed of 10 packets per second, and the receiving device only has one receiving antenna to receive the WiFi data packets.

[0040] S2: Calculate the energy spectrum of the CSI signal of each subcarrier. Specifically, according to the data packets received by each receiving antenna, obtain the channel state information CSI of all 52 subcarriers accordingly, which are 52 complex numbers, and calculate the energy spectrum of the CSI signal on each subcarrier, that is, the square of the amplitude of the complex CSI signal.

[0041] S3: Obtain the variance of the CSI energy spectrum of each subcarrier according to the CSI signal energy spectrum within a period window. Specifically, according to the energy spectrum of the CSI signal within a period window of 2 seconds, that is, the CSI signal energy spectrum obtained from 20 data packets, obtain a 52X20 CSI energy spectrum matrix Calculate the variance of the CSI energy spectrum of each subcarrier, which is a 52X1 variance vector.

[0042] S4: Judge whether the mean value of the obtained variance of the CSI energy spectrum is higher than the preset body movement threshold d1. Specifically, judge whether the mean value of the variance vectors of the CSI energy spectra of all subcarriers obtained is higher than the preset body movement threshold d1. When the judgment result is yes, it is determined that there is someone in the car, and the operation process can be ended; if the judgment result is no, then continue with step S5.

[0043] S5: Process the CSI energy spectrum according to a preset method to obtain the corresponding respiratory signal spectrum. Specifically, when the judgment result is negative, process the energy spectrum signal of the CSI within 10 seconds for each subcarrier, that is, perform a 128-point fast Fourier transform on each row of a 52×100 CSI energy spectrum matrix to obtain the corresponding respiratory signal spectrum, calculate its amplitude, and obtain a 52×128 matrix in the frequency domain

[0044] S6: Determine whether the peak point of the obtained respiratory signal spectrum is within the human respiratory frequency range. Specifically, the normal human respiratory frequency f ranges from 0.15 Hz to 0.50 Hz. When corresponding to the frequency domain of the FFT, the calculation formula for the interval of the corresponding frequency point K is that is, K ∈ [3, 7]. For the respiratory signal spectrum of each subcarrier, that is, each row of the matrix , find the frequency point corresponding to the maximum amplitude value, take the mode of the peak points of the respiratory signal spectra of all subcarriers, and determine whether it is within the interval [3, 7]. When the judgment result is positive, it is determined that there is someone in the car; when the judgment result is negative, it is determined that there is no one in the car

[0045] The above are some specific implementation manners of the present invention, but the present invention is not limited to the above manners. All simple transformations of the technical features of the present invention, and all equivalent changes or modifications made according to the structure, features, and principles described in the scope of the patent application of the present invention, will fall within the protection scope of the present invention

Claims

1. A method for human presence perception based on in-vehicle WiFi, characterized in that Including: Step 1: A receiving device receives a radio frequency signal emitted from a transmitting device, and the receiving device has a receiving antenna; Step 2: Obtain the channel state information CSI of multiple subcarriers based on the radio frequency signals received by each receiving antenna, and calculate the energy spectrum of the CSI signal on each subcarrier; Step 3: Calculate the variance of the CSI energy spectrum of each subcarrier according to the energy spectrum of the CSI signal within a time window; Step 4: Determine whether the mean value of the CSI energy spectrum variances of all obtained subcarriers is higher than a preset body movement threshold. When the determination result is yes, it is determined that there is someone in the car, and go to Step 6; When the determination result is no, process the energy spectrum signal of the CSI of each subcarrier according to a preset method to obtain a respiration signal spectrum; Step 5: Determine whether the peak point of the obtained respiration signal spectrum is within the human respiration frequency range. When the determination result is yes, it is determined that there is someone in the car; when the determination result is no, it is determined that there is no one in the car; Step 6: End.

2. The human presence sensing method according to claim 1, wherein The radio frequency signal is a WiFi data packet emitted by the same transmitting device, and the transmitting device includes an in-vehicle WiFi device.

3. The human presence sensing method according to claim 1, characterized in that The calculation of the energy spectrum of the CSI signal on each subcarrier includes: Set the system sampling rate to F s , with the unit of Hz. Then, the channel state information on the subcarrier f at time t is calculated as H(t,f), and the energy spectrum G(t,f) is calculated according to the following formula: G(t,f) = |H(t,f)| 2 , where H(t,f) is a complex number.

4. The human presence sensing method according to claim 3, wherein The calculation of the variance of the CSI energy spectrum of each subcarrier includes: Set the size of a time window to be W1, with the unit of seconds. The channel state information within the time window is all the sampling points within this time window, which are the sampling points from t1 seconds to t2 seconds. For each sub - carrier, calculate the variance of the CSI energy spectrum Var(f) within its corresponding time window according to the following formula: Among them is the mean value of the CSI energy spectrum on subcarrier f within the time window.

5. The human presence sensing method according to claim 4, characterized in that When comparing the mean value of the CSI energy spectrum variances of all subcarriers with a preset body movement threshold d1, when the determination result is no, it further includes: updating the preset body movement threshold d1.

6. The human presence sensing method according to claim 5, wherein, The updating of the preset body movement threshold d1 includes: Set a time window W2 in seconds, and determine whether the standard deviation of the mean value of the energy spectrum variances of all subcarriers within the time window W2 is higher than an update threshold d2. When the determination result is yes, do not update the body movement threshold d1; when the determination result is no, update it to a new body movement threshold d1.

7. The human presence sensing method according to claim 6, wherein The new body movement threshold d1 is the sum of the mean value of the CSI energy spectrum variances of all subcarriers within the time window W2 plus three times its standard deviation.

8. The human presence sensing method according to claim 7, wherein The processing of the energy spectrum signal of the CSI of each subcarrier according to a preset method to obtain a respiration signal spectrum further includes: converting the time-domain CSI energy spectrum signal within a time window W3 into a corresponding frequency-domain signal through fast Fourier transform FFT, and finding the frequency point corresponding to its maximum amplitude. The unit of the time window W3 is seconds.

9. The human presence sensing method according to claim 8, characterized in that, The determination of whether the peak point of the obtained respiration signal spectrum is within the human respiration frequency range further includes: calculating the corresponding frequency according to the frequency point corresponding to the found maximum amplitude; if the frequency is within the interval of the normal human respiration frequency range, it is determined that there is someone in the car, where the normal human respiration frequency range is 0.15Hz - 0.50Hz.

10. A human presence perception system based on in-vehicle WiFi, characterized in that Including: A transmitting device, a receiving device, a processor, a memory, and a controller. The controller controls the transmitting device and the receiving device to transmit and receive radio frequency signals. The memory stores a medium with program code. When the processor reads the program code stored in the medium, the system can execute the method according to any one of claims 1 - 9.

Images