LoRa Wireless Multi-Person Sensing Method Based on Signal Vector Separation

Abstract

The present invention relates to a LoRa wireless multi-person breathing perception method based on signal vector separation, which solves the problem that it is difficult to effectively separate the breathing signals of multiple targets in the field of wireless perception, so as to realize wireless breathing perception using LoRa devices in a multi-person scenario. First, a single-transmission and dual-reception LoRa device is used to collect the breathing data of the target; then the data is preprocessed, including removing phase offset using the signal ratio, smoothing filtering, etc.; secondly, the signal is re-segmented, modeled as a form of a static signal plus multiple dynamic signals, and according to the fact that the breathing movement of the human body can be approximated as a sinusoidal movement in a steady state, parameters are introduced to describe each dynamic signal in more detail; finally, the particle swarm optimization algorithm is used to optimize and solve each parameter, and the breathing information of each target is obtained thereby.

Description

Technical Field

[0001] The present invention belongs to the field of wireless sensing, and specifically relates to a method for separating the breathing signals of multiple targets in a vector of LoRa reflected signals to achieve simultaneous breathing perception of multiple people. Background Art

[0002] In recent years, with the development of Internet of Things technology, wireless sensing technology has become increasingly popular, such as WiFi, RFID, millimeter wave radar, visible light, etc. Compared with traditional wearable sensing, wireless sensing increases the comfort and convenience of users without adding extra wires to the sensing device. However, although most current wireless sensing devices use both direct paths and reflected paths for communication, they can only sense using weak target reflected signals, so their sensing distances are very small.

[0003] As a new wireless technology, LoRa uses spread spectrum modulation to obtain coding gain through spread spectrum modulation and can receive signals with a lower signal-to-noise ratio. Therefore, the communication distance of LoRa can reach several kilometers. Thus, as a wireless sensing device, the sensing distance of LoRa is much greater than that of other wireless sensing devices such as WiFi. When using a LoRa device for sensing, the received signal can be divided into a static signal and a dynamic signal. The static signal is caused by signals reflected by static objects in the environment, while the dynamic signal is caused by moving sensing targets. When the target moves, the phase and amplitude of the dynamic signal will change accordingly, so that the dynamic signal vector will rotate around the static signal vector in the complex plane.

[0004] However, due to carrier frequency offset (CFO) and sampling frequency offset (SFO) during signal transmission and sampling, random phase offsets will be caused, resulting in the destruction of the phase of the received signal. Therefore, the received signal cannot be directly used for sensing. And LoRa gateways usually have two receiving antennas, and these two antennas share a clock, so they have the same phase offset caused by CFO and SFO. Therefore, the random phase offset can be eliminated by taking the ratio of the two signals received by the two antennas. At the same time, according to the Möbius transformation, the result of the signal ratio will only rotate and translate the signal in the vector space, without changing its shape and direction. Therefore, the result of the signal ratio can still use the traditional signal vector model. Summary of the Invention

[0005] Technical Problems to be Solved

[0006] To avoid the deficiencies of the prior art, the present invention provides a LoRa wireless multi-person respiration sensing method based on signal vector separation to solve the problem that when multiple targets are in the same scenario, traditional wireless respiration sensing methods cannot effectively separate the respiration sensing data of multiple targets.

[0007] Technical solution

[0008] A LoRa wireless multi-person respiration sensing method based on signal vector separation uses a single-transmission and dual-reception LoRa device to achieve wireless respiration sensing of multiple targets in a multi-person scenario; its characteristics are as follows:

[0009] Step 1: Take the ratio of the signals received by the two LoRa antennas to remove the random phase offset caused by the carrier frequency offset CFO and the sampling frequency offset SFO. The signal ratio uses the traditional signal vector model:

[0010]

[0011] Among them, H(t) represents the ratio of the signals of the two receiving antennas, e φoffset represents the random phase offset of the signal, and H1 and H2 respectively represent the signals received by the two signal receiving antennas after removing the random phase offset. According to the Möbius transformation, the signal ratio H(t) can be used for wireless sensing;

[0012] Step 2: Model the signal ratio after removing the random phase offset as the form of an environmental static signal plus two dynamic signals, and the two dynamic signals are respectively caused by the breathing movements of the two targets:

[0013] H(t) = H s (t) + H d1 (t) + H d2 (t)

[0014] Among them, H s (t) represents the environmental static signal, and H d1 (t), H d2 (t) respectively represent the dynamic signals caused by the breathing movements of two human targets;

[0015] Step 3: Introduce two parameters δ 01 , ω1 to describe the phase of the breathing movement of target one. Among them, δ 01 represents the phase value of the breathing at the initial moment of monitoring, and ω1 represents the angular velocity of the breathing. Then the phase can be expressed as:

[0016]

[0017] Step 4: Introduce parameters α1 and β1 to describe the phase θ1 of the breathing dynamic signal of Target 1, where α1 represents the minimum phase value of the dynamic signal and β1 represents the magnitude of the phase change range of the dynamic signal; then the dynamic signal θ1 can be expressed as:

[0018]

[0019] Step 5: Introduce parameter A1 to represent the amplitude of the breathing dynamic signal of Target 1, then the breathing dynamic signal H of Target 1 d1 can be expressed as:

[0020]

[0021] Introduce parameter A2 to represent the amplitude of the breathing dynamic signal of Target 2, then the breathing dynamic signal H of Target 2 d2 can be expressed as:

[0022]

[0023] Step 6: The received signal H(t) can be expressed as:

[0024]

[0025] Step 7: Use the LoRa signal sending node and the USRP B210 gateway to transmit and receive signals. In the current scenario without test targets, turn on the device in advance to measure the environmental static signal H s (t). After the static signal measurement is completed, let the two target persons being measured sit still and keep a steady breathing state within the sensing range of the LoRa signal facing the transceiver antenna to start receiving induction data;

[0026] Step 8: Take the ratio of the two received signals to remove the phase offset, and use the Savitzky-Golay filter to perform filtering and smoothing processing on it. Set a sliding time window of appropriate size for the data, and use the particle swarm optimization algorithm to optimize and solve each unknown parameter. According to the solved angular velocity ω, the breathing frequency of the corresponding target can be obtained; considering time efficiency, within a short time, the parameters A1, α1, β1, A2, α2, β2 can be regarded as constant values. Therefore, the values of these parameters can be extended from the optimization results of the previous sliding time window, and then optimized and updated after a long time, so that only δ 01 , ω1, δ 02 , ω2 need to be optimized and solved;

[0027] Step 9: When the number of people in the scenario is greater than two, the number of dynamic signals in the model can be added according to the target number of people in the scenario, so that each target corresponds to a breathing dynamic signal, and wireless breathing perception of multiple people can be achieved.

[0028] In step 7, the sensing range of the LoRa signal is 8 - 20m.

[0029] A computer system, comprising: one or more processors, a computer-readable storage medium for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the above method.

[0030] A computer-readable storage medium, characterized in that it stores computer-executable instructions, and the instructions are used to implement the above method when executed.

[0031] Beneficial effects

[0032] A LoRa wireless multi-person perception method based on signal vector separation provided by the present invention first uses a single-transmission and dual-reception LoRa device to collect breathing data of targets; then preprocesses the data, including removing phase offset using signal ratio, smoothing filtering, etc.; secondly, re-segments the signal, models it in the form of a static signal plus multiple dynamic signals, and according to the breathing movement of the human body in a stable state can be approximated as a sine movement, introduces parameters to describe each dynamic signal in more detail; finally, uses the particle swarm optimization algorithm to optimize and solve each parameter, and obtains the breathing information of each target accordingly. The method of the present invention can effectively separate the breathing signals of multiple targets in a multi-person scenario using a LoRa device with two receiving antennas, so as to realize wireless breathing monitoring of multiple targets simultaneously. Description of the drawings

[0033] The drawings are only for the purpose of showing specific embodiments, and are not considered to be a limitation of the present invention. Throughout the drawings, the same reference signs represent the same components.

[0034] Figure 1 It is a method flow framework diagram in the present invention;

[0035] Figure 2 It is a mapping schematic diagram of the breathing of the target human body to the change of dynamic signals in the present invention;

[0036] Figure 3 It is an example deployment diagram of the transceiver antennas in the present invention.

[0037] Figure 4 It is an example diagram of LoRa nodes and USRP gateways in the present invention. Detailed implementation manners

[0038] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0039] The present invention utilizes the following principle: Considering the two-person scenario, after obtaining the signal ratio of the received signal to remove the random phase offset, it is modeled as a form of an environmental static signal plus two dynamic signals, and the two dynamic signals are respectively caused by the breathing movements of the two targets. Among them, when the environment remains unchanged, the environmental static signal can be regarded as a fixed value, so the value of the environmental static signal can be measured in advance. Based on existing research, when the human body is in a stable state, the breathing movement of the human body can be regarded as a sinusoidal movement pattern. Since the change of the dynamic signal is caused by the breathing movement of the human body, when the human body performs a periodic sinusoidal breathing movement, in the complex plane, the dynamic signal will also show a corresponding periodic sinusoidal back-and-forth swing. In addition, since the amplitude of the human breathing movement is relatively small, generally 5 to 12 mm, the amplitude change of the dynamic signal caused by it is very small, so we can regard the amplitude of the dynamic signal as a fixed value. Then in the two-person scenario, for the received signal, we can decompose it into a static signal and two dynamic signals with fixed amplitudes that swing back and forth sinusoidally around the static signal. For each dynamic signal, we introduce unknown parameters such as the target breathing angular velocity, the target breathing initial phase, the dynamic signal amplitude, and the maximum and minimum values of the dynamic signal phase, and solve these parameters through the particle swarm optimization algorithm according to the collected data. Finally, according to these parameters, the breathing state information of each target can be obtained.

[0040] Refer to Figure 1 , the specific implementation steps of the present invention are as follows:

[0041] Step 1: Take the ratio of the signals received by the two LoRa antennas to remove the random phase offset caused by CFO and SFO. According to the Möbius transformation, the result of the signal ratio will only rotate and translate the signal in the vector space, without changing its shape and direction, so the result of the signal ratio can still use the traditional signal vector model.

[0042]

[0043] Step 2: Model the signal ratio after removing the random phase offset as a form of an environmental static signal plus two dynamic signals, and the two dynamic signals are respectively caused by the breathing movements of the two targets.

[0044] H(t) = H s (t) + H d1 (t) + H d2 (t)

[0045] Step 3: First, model the breathing pattern of Target 1 among the two targets. Since the breathing movement of the human body in a stable state can be regarded as a sinusoidal movement, two parameters δ 01 and ω1 are introduced to describe the phase of the breathing movement of Target 1. Among them, δ 01 represents the phase value of breathing at the initial moment of monitoring, and ω1 represents the angular velocity of breathing. Then the phase can be expressed as:

[0046]

[0047] Step 4: The breathing movement of the target causes changes in the dynamic signal. Since the amplitude of the breathing movement is small, the amplitude change of the dynamic signal is small and can be regarded as a constant value. And the phase of the dynamic signal is associated with the phase of the breathing movement, and it changes sinusoidally in synchronization with the breathing movement. Parameters α1 and β1 can be introduced to describe the phase θ1 of the breathing dynamic signal of Target 1. Among them, α1 represents the minimum phase value of the dynamic signal, and β1 represents the size of the phase change range of the dynamic signal. Then the dynamic signal θ1 can be expressed as:

[0048]

[0049] Step 5: Introduce the parameter A1 to represent the amplitude of the breathing dynamic signal of Target 1. Then the breathing dynamic signal H d1 of Target 1 can be expressed as:

[0050]

[0051] Step 6: Similarly, the breathing dynamic signal H d2 of Target 2 can also be expressed in the same form. Then, for the received signal H(t), it can be expressed as:

[0052]

[0053] Step 7: Use the LoRa signal sending node and the USRP B210 gateway in Figure 4 to perform signal transmission and reception, and deploy the signal transmission and reception antennas according to Figure 3 ( Figure 4 shows the LoRa signal sending node and the USRP B210 gateway respectively. Among them, the LoRa signal sending node can continuously send LoRa signals, which are received by the USRP B210 gateway. Connect the LoRa signal sending node withFigure 3 is connected to one of the directional antennas, and the USRP B210 gateway is connected to Figure 3 the other two directional antennas in it, and the three directional antennas are placed side by side as Figure 3 shown. After the arrangement, turn on the LoRa signal sending node and the USRP B210 gateway to start collecting data.). In the current scenario, when there is no test target, turn on the device in advance to measure the static signal H s (t) of the environment. After the static signal measurement is completed, let the two tested target persons sit still within 8m of the transceiver antenna and maintain a steady breathing state to start receiving induction data.

[0054] Step Eight: Remove the phase offset by taking the ratio of the two groups of received signals, and use the Savitzky-Golay filter to filter and smooth them. Set a sliding time window of appropriate size for the data, and use the particle swarm optimization algorithm to optimize and solve each unknown parameter. According to the solved angular velocity ω, the breathing frequency of the corresponding target can be obtained. Considering time efficiency, within a short time, the parameters A1, α1, β1, A2, α2, β2 can be regarded as fixed values. Therefore, the values of these parameters can be extended from the optimization results of the previous sliding time window. After a long time, optimize and update them, so that only δ 01 , ω1, δ 02 , ω2 need to be optimized and solved in the current sliding time window.

[0055] Step Nine: When the number of people in the scenario is greater than two, the number of dynamic signals in the model can be added according to the number of targets in the scenario, so that each target corresponds to a breathing dynamic signal, and wireless breathing perception of multiple people can be realized.

[0056] The present invention is a method for multi-person breathing perception in the LoRa wireless sensing field. Using the method of signal vector separation, the breathing signals of each target in a multi-target scenario are effectively separated, thus realizing the real-time perception of the breathing state of each target. In addition to LoRa, the present invention is also applicable to other wireless sensing devices such as WiFi.

[0057] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should be covered within the protection scope of the present invention.

Claims

1. A LoRa wireless multi - person breathing perception method based on signal vector separation, which uses a single - transmit and dual - receive LoRa device to achieve wireless breathing perception of multiple targets in a multi - person scenario; characterized in that The steps are as follows: Step 1: Calculate the ratio of the signals received by the two LoRa antennas to remove the random phase offset caused by the carrier frequency offset (CFO) and the sampling frequency offset (SFO). The signal ratio uses the traditional signal vector model: Among them, H(t) represents the ratio of the signals of two receiving antennas, and e φoffset represents the random phase offset of the signal. H1 and H2 respectively represent the signals received by the two signal receiving antennas after removing the random phase offset. According to the Möbius transformation, the signal ratio H(t) can be used for wireless sensing; Step 2: Model the signal ratio after removing the random phase offset as the form of an ambient static signal plus two dynamic signals, where the two dynamic signals are respectively caused by the breathing movements of two targets: H(t) = H s (t) + H d1 (t) + H d2 (t) Among them, H s (t) represents the environmental static signal, and H d1 (t), G d2 (t) respectively represent the dynamic signals caused by the breathing movements of two human targets; Step 3: Introduce two parameters δ 01 , ω1 to describe the phase of the target first respiratory movement, where δ 01 represents the phase value of respiration at the initial monitoring moment, ω1 represents the angular velocity of respiration, then the phase can be expressed as: Step 4: Introduce parameters α1 and β1 to describe the phase θ1 of the breathing dynamic signal of Target 1, where α1 represents the minimum phase value of the dynamic signal and β1 represents the magnitude of the phase change range of the dynamic signal. Then the dynamic signal θ1 can be expressed as: Step 5: Introduce parameter A1 to represent the amplitude of the respiratory dynamic signal of Target 1, then the respiratory dynamic signal H of Target 1 d1 can be expressed as: The introduced parameter A2 represents the amplitude of the respiratory dynamic signal of Target 2, and the respiratory dynamic signal H of Target 2 d2 can be expressed as: Step 6: The received signal H(t) can be expressed as: Step 7: Use the LoRa signal sending node and the USRP B210 gateway to send and receive signals. When there is no test target in the current scenario, turn on the device in advance to measure the ambient static signal H s (t). After the static signal measurement is completed, let the two target persons to be measured sit still within the sensing range of the LoRa signal facing the transceiver antenna and maintain a steady breathing state to start receiving the induction data; Step 8: Remove the phase offset by taking the ratio of the two sets of received signals, perform filtering and smoothing on them using the Savitzky-Golay filter, set a sliding time window of appropriate size for the data, use the particle swarm optimization algorithm to optimize and solve for each unknown parameter, and obtain the breathing frequency of the corresponding target based on the solved angular velocity ω; Considering time efficiency, within a short period of time, the parameters A1, α1, β1, A2, α2, β2 can be regarded as fixed values, so the values of these parameters can use the optimization results of the previous sliding time window, and optimize and update them after a long period of time, so that only δ 01 , ω1, δ 02 , ω2 need to be optimized and solved; Step 9: When the number of people in the scenario is greater than two, the number of dynamic signals in the model can be added according to the number of targets in the scenario, so that each target corresponds to a breathing dynamic signal, and wireless breathing perception of multiple people can be achieved.

2. The LoRa wireless multi - person breathing perception method based on signal vector separation according to claim 1, characterized in that In Step 7, the sensing range of the LoRa signal is 8 - 20 m.

3. A computer system, characterized in that Comprising: One or more processors, a computer-readable storage medium for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method according to Claim 1.

4. A computer - readable storage medium, characterized in that Stored with computer-executable instructions, the instructions being used to implement the method according to Claim 1 when executed.

Images