A user position positioning method based on human body breathing Wi-Fi reflection signal

Abstract

The application discloses a user position positioning method based on human body breathing Wi-Fi reflection signals and belongs to the technical field of Wi-Fi breathing monitoring and human body passive positioning. Firstly, the application eliminates CFO and SFO phase noise in CSI by synthesizing reference signals through multiple antenna signals and comparison; then, the application filters the signals and extracts time domain optimal projection angles representing breathing movement; further, according to the linear variation law of the projection angles with subcarrier and antenna index, the application respectively analyzes the time difference of flight and the angle of arrival information of the breathing reflection path; finally, through the construction of a position grid, the application compensates and matches the measured projection angles by using theoretical geometric parameters, and realizes user position positioning according to the criterion that the overall variance of the compensated projection angles is minimum. The application robustly extracts positioning parameters by using the projection angle variation law of the breathing signals, effectively overcomes phase noise interference, and realizes high-robustness and high-precision non-contact passive personnel positioning.

Description

Technical Field

[0001] This invention relates to the field of Wi-Fi breathing monitoring and passive human positioning technology, specifically a user location positioning method based on human breathing Wi-Fi reflection signals. It utilizes equivalent signal analysis to obtain the signal flight time and angle of arrival, and then solves the absolute position of a static user. Background Technology

[0002] With the rapid development of IoT technology and the widespread adoption of Wi-Fi devices, Wi-Fi-based passive positioning technology has attracted widespread attention due to its advantages such as low deployment cost and wide applicability. Traditional Wi-Fi passive positioning technology usually requires the user to be in motion, and it achieves positioning by analyzing signal changes caused by the user's movement (such as Doppler frequency shift). However, in scenarios such as indoor offices and home rest, users are mostly in a static state, making this type of technology difficult to be effective.

[0003] In existing static positioning solutions, camera-based methods suffer from privacy risks and high deployment costs, making them unsuitable for private spaces. Methods based on traditional Wi-Fi signal characteristics (such as received signal strength RSS) are significantly affected by multipath effects and environmental interference, resulting in limited positioning accuracy. Furthermore, existing Wi-Fi positioning technologies have the following drawbacks: first, they rely on high data packet rates (typically around 1000Hz), consuming additional channel resources and reducing normal communication efficiency; second, some methods obtain location by integrating user velocity, which can easily lead to error accumulation; and third, they cannot effectively distinguish between the reflected signals of a static human body and static objects such as walls and furniture, making it difficult to achieve accurate positioning of static users.

[0004] Although existing studies have verified the feasibility of using Wi-Fi signals to detect respiratory cycles, the impact of respiratory signals on Wi-Fi signals is far weaker than the signal changes caused by human movement. Furthermore, traditional signal separation methods easily introduce residual signal components, making location feature extraction difficult. In addition, issues in CSI measurements such as carrier frequency offset (CFO), sampling frequency offset (SFO), and initial phase offset (IPO) disrupt the theoretical relationship between signal features and location required for positioning, further increasing the difficulty of extracting location information from respiratory signals. Therefore, developing a method capable of accurately extracting location features from static user respiratory signals at low data packet rates, achieving calibration-free, high-precision static positioning, is of significant practical importance. Summary of the Invention

[0005] The purpose of this invention is to propose a user location positioning method based on human breathing Wi-Fi reflection signals to solve the problem that existing technologies cannot locate stationary targets. This invention utilizes the periodic signal characteristics generated by the micro-movement of the chest cavity during human breathing, and through the "equivalent signal analysis" method, accurately extracts location features without completely separating weak dynamic signals, thereby achieving high-precision static positioning of users who do not carry devices.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A user location positioning method based on human respiratory Wi-Fi reflection signals includes the following steps: S1. Signal Acquisition and Cleaning: The receiver acquires the raw signal containing the dynamic changes in Channel State Information (CSI) caused by human respiration through multiple antennas; the raw signal is cleaned to eliminate carrier frequency offset (CFO) and sampling frequency offset (SFO) to obtain the cleaned Channel State Information (CSI) ratio signal. S2. Time Domain Analysis: Bandpass filtering is performed on the cleaned Channel State Information (CSI) ratio signal to extract the signal component corresponding to the respiratory frequency; the optimal projection axis that maximizes the respiratory periodicity is determined from the signal component, and the corresponding optimal projection angle sequence is calculated; S3. Frequency Domain and Spatial Domain Analysis: Based on the optimal projection angle sequence, the time difference of flight (TDoF) between the dynamic path and the static path is extracted by analyzing the relationship between the optimal projection angle and the subcarrier index; the angle of arrival (AoA) corresponding to the breathing signal is extracted by analyzing the relationship between the optimal projection angle and the receiving antenna index; combined with the known transceiver position information, the geometric information of the candidate reflection path corresponding to the breathing signal is calculated based on the time difference of flight (TDoF) between the dynamic path and the static path and the signal angle of arrival (AoA); S4. Location Positioning: Divide the positioning area into grids; for each candidate grid location, calculate the time difference of flight (TDoF) between the theoretical dynamic path and the static path, as well as the theoretical angle of arrival (AoA), based on their geometric relationships; use the time difference of flight (TDoF) between the theoretical dynamic path and the static path, and the theoretical angle of arrival (AoA) to compensate for the measured optimal projection angle sequence, and calculate the overall variance of the compensated projection angle on all subcarriers and all antennas; traverse all grids, and determine the grid location that minimizes the overall variance as the user's estimated location.

[0008] Preferably, the signal cleaning of the original signal in S1 specifically includes: S11. Reference signal synthesis: The raw channel state information (CSI) signals collected by multiple antennas of the receiver are weighted and summed to synthesize a reference signal. The weights are selected according to the criterion of maximizing the ratio of the signal average to the variance in order to suppress the breathing dynamic component in the reference signal. S12. Phase Error Elimination: Divide the original Channel State Information (CSI) signals received by each antenna by the synthesized reference signal to obtain the Channel State Information (CSI) ratio signal, so as to eliminate the common carrier frequency offset (CFO) and sampling frequency offset (SFO) phase errors between different antennas.

[0009] Preferably, determining the optimal projection axis from the signal components in step S2 specifically involves: The filtered respiratory signal is mapped onto the complex plane. By calculating the variance or respiratory signal-to-noise ratio of the signal in different projection axis directions, the axis that maximizes the variance or respiratory signal-to-noise ratio is selected as the optimal projection axis. The angle between the optimal projection axis and the reference axis constitutes the optimal projection angle sequence.

[0010] Preferably, S2 further includes using a complex smoothing algorithm to correct the optimal projection angle sequence in order to solve the phase jump problem caused by projection blur.

[0011] Preferably, the step S3, which involves analyzing the relationship between the optimal projection angle and the subcarrier index to extract the time difference of flight (TDoF) between the dynamic path and the static path, specifically involves linearly fitting the optimal projection angles corresponding to different subcarrier indices at the same time. The slope of the fitted line is proportional to the time difference of flight (TDoF) between the dynamic path and the static path.

[0012] Preferably, the extraction of the signal angle of arrival (AoA) by analyzing the relationship between the optimal projection angle and the receiving antenna index in step S3 specifically includes: S31. Obtain the optimal projection angle difference value of each receiving antenna relative to the reference antenna to form a spatial phase difference sequence; S32. Using the known geometric positional relationship of the transceiver, calculate and eliminate the influence of the initial phase offset (IPO) caused by hardware on the spatial phase difference sequence; S33. Based on the geometric relationship between the spatial phase difference sequence after eliminating the initial phase offset (IPO) and the antenna spacing, the signal arrival angle (AoA) corresponding to the breathing signal is calculated.

[0013] Preferably, the compensation of the measured optimal projection angle sequence using the time difference of flight (TDoF) between the theoretical dynamic path and the static path and the theoretical angle of arrival (AoA) in S4 is specifically as follows: for each subcarrier and each antenna, a theoretical phase compensation amount is calculated based on the time difference of flight (TDoF) between the theoretical dynamic path and the static path and the theoretical angle of arrival (AoA); the measured optimal projection angle is subtracted from the corresponding theoretical phase compensation amount to obtain the compensated projection angle.

[0014] Compared with the prior art, the present invention has the following beneficial effects: (1) First realization of passive Wi-Fi positioning for static users: This invention utilizes the inherent physiological characteristic of human breathing to effectively distinguish between the reflected signals of static human bodies and static objects such as walls and furniture, filling the gap in existing technologies that cannot locate static users and expanding the application scenarios of Wi-Fi positioning.

[0015] (2) Low data packet rate requirement, saving channel resources: The present invention only requires a data packet rate of 50 Hz or higher to achieve meter-level positioning accuracy. Compared with the existing technology's data packet rate requirement of 1000 Hz, it greatly reduces the occupation of channel resources and ensures normal Wi-Fi communication efficiency.

[0016] (3) Calibration-free design reduces deployment costs: This invention does not require a complex environmental calibration process and a large amount of measured data training. By combining theoretical models with equivalent signal analysis, it directly extracts location features from CSI data, reducing labor and time costs and improving the system's versatility and deployment convenience.

[0017] (4) High positioning accuracy and no error accumulation: This invention directly solves the absolute position by extracting the signal time of flight (ToF) and the signal angle of arrival (AoA), avoiding the error accumulation problem caused by the positioning method based on velocity integration. It can achieve a positioning accuracy of median 0.89 m in both open spaces and office scenarios.

[0018] (5) Strong anti-interference capability: This invention effectively suppresses the influence of environmental noise, multipath effect and phase shift on positioning through reference signal synthesis, multi-stage phase offset clearing and projection angle correction, thereby improving the stability and reliability of the system in complex environments. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings involved in the embodiments are now briefly described. Obviously, the drawings in the following description are merely illustrative of some embodiments of the present invention. For those skilled in the art, other forms of drawings can be constructed based on these drawings without creative effort.

[0020] Figure 1 This is an overall flowchart of a user location positioning method based on human respiratory Wi-Fi reflection signals proposed in this invention; Figure 2 This is a schematic diagram illustrating the influence of respiration on the complex frequency domain signal changes mentioned in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of frequency domain and spatial domain analysis mentioned in Embodiment 2 of the present invention; Figure 4 This is a diagram illustrating the fusion positioning effect of Angle of Arrival (AOA) and Time of Flight (ToF) mentioned in Embodiment 2 of the present invention. Detailed Implementation

[0021] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0022] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0023] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0024] This invention proposes a user location positioning method based on Wi-Fi reflection signals from human respiration. By utilizing the periodic signal characteristics generated by the micro-movement of the chest cavity during human respiration, and through the "equivalent signal analysis" method, the location features are accurately extracted without completely separating the weak dynamic signals, thereby achieving high-precision static positioning of users who do not carry devices.

[0025] The following explains the technical terms mentioned in this invention: Wi-Fi respiratory monitoring: This refers to a technology that uses the propagation characteristics of Wi-Fi signals to detect signal changes caused by the periodic rise and fall of the chest during human respiration, thereby enabling the monitoring of human respiratory status without requiring the user to wear any sensing devices.

[0026] Wi-Fi passive positioning: a technology that tracks a user's location by analyzing the characteristics of Wi-Fi signals reflected from the human body (such as phase and amplitude changes) without requiring the user to carry intrusive devices such as mobile phones or sensors.

[0027] Channel State Information (CSI) is used to characterize the impact of the wireless channel on the amplitude and phase of the signal. It is a superposition of the direct path signal and multipath reflected signals (including reflected signals from people, walls, furniture, etc.), and contains rich information about the signal propagation process.

[0028] Time of Flight (ToF): This refers to the time it takes for a Wi-Fi signal to travel from the transmitter to the receiver. The signal propagation distance can be calculated based on the speed of electromagnetic wave propagation, and it is one of the key parameters for positioning.

[0029] Angle of Arrival (AoA): This refers to the angle between the Wi-Fi signal and the normal direction of the antenna array when the signal arrives at the receiver's antenna array. Angle estimation can be achieved by receiving data from multiple antennas, providing directional information for positioning.

[0030] Based on the above, the following describes a user location positioning method based on human respiratory Wi-Fi reflection signals proposed in this invention, with specific examples.

[0031] Example 1: This invention proposes a user location positioning method based on human respiratory Wi-Fi reflection signals. Employing the core concept of "equivalent signal analysis," it transforms the weak respiratory signal characteristics into variations in the projection angle. The specific process includes four stages: signal cleaning, time-domain analysis, frequency / spatial-domain analysis, and location positioning. The details are as follows: 1. Signal cleaning phase This phase aims to remove phase noise (CFO and SFO) from CSI measurements while preserving projection rules that imply location information.

[0032] 1) Reference Signal Synthesis: To construct a high-quality CSI ratio (CSIRatio) denominator, this invention proposes a reference signal synthesis mechanism. A "reference signal" is synthesized by weighted summation of the original signals from multiple receiver antennas. The weights are selected to maximize the "static-dynamic energy ratio" (i.e., the ratio of the mean to the variance), thereby suppressing the dynamic breathing component in the reference signal and approximating it as a purely static signal.

[0033] 2) Carrier Frequency Offset (CFO) and Sampling Frequency Offset (SFO) Removal: Using the synthesized reference signal as the denominator, the original CSI signals of each antenna are divided by this reference signal. Since CFO and SFO are the same across different antennas, these two time-varying and frequency-varying phase errors can be eliminated through the division operation, retaining only the Initial Phase Offset (IPO) and the dynamic components containing position information.

[0034] 2. Time Domain Analysis This stage aims to extract the "optimal projection angle" that can characterize respiratory movements from the cleaned signals.

[0035] 1) Respiratory signal filtering: Considering the normal respiratory frequency range (approximately 0.17 Hz-0.62 Hz), the CSI ratio signal is bandpass filtered to remove high-frequency noise and other motion interference.

[0036] 2) Optimal projection angle selection: The respiratory signal on the complex plane is considered as an arc. By calculating the signal variance or respiratory signal-to-noise ratio (BNR) on different projection axes, the axis that maximizes the signal periodicity is selected as the "optimal projection axis". The angle between this projection axis and the horizontal axis is the "optimal projection angle". To solve the phase jump caused by projection blur, a complex smoothing algorithm is used to correct the projection angle sequence.

[0037] 3. Frequency domain and spatial domain analysis This invention utilizes the "projection rule," which states that the ToF and AoA characteristics of the respiratory signal are implicit in the rate of change of the optimal projection angle with respect to the subcarrier and the antenna, respectively.

[0038] 1) Multicarrier Analysis (Time-of-Flight Extraction): According to the equivalent analysis theory, the Time-of-Flight (ToF) of a dynamic signal is a linear change in the optimal projection angle with the subcarrier index. By calculating the slope of the projection angle as a function of the subcarrier, the Time-of-Flight (TDoF) between the dynamic and static paths can be determined. Since the transceiver positions are known, the ToF of the direct wave can be calculated, and thus the ToF of the human body reflection path can be solved.

[0039] 2) Multi-antenna analysis (AoA extraction): Similar to multi-carrier analysis, the AoA of dynamic signals also manifests as the phase change of the optimal projection angle with the antenna index.

[0040] 3) Initial Phase Offset (IPO) Removal: For the initial phase offset (IPO) existing in the spatial domain, the time-invariant characteristics of the IPO and the known geometric relationship of the transceiver position are used to estimate and remove the IPO by comparing the theoretical phase difference of the direct wave with the measured phase difference, thereby restoring the true AoA information.

[0041] 4. Location positioning Location Matching: The positioning area is divided into a grid. For each candidate location, the theoretical Time-of-Flight (ToF) and AoA (AoA) are calculated based on geometric relationships. These theoretical values ​​are used to compensate for the measured projection angle sequence. If the candidate location is the true location, the compensated projection angle should converge across all subcarriers and antennas, i.e., the variance should be minimized. The point with the minimum variance is found by traversing all grids; this is the estimated location of the user.

[0042] Example 2: Based on Example 1, but with some differences, please refer to Figure 1-4 The user location positioning method based on human respiratory Wi-Fi reflection signals proposed in this invention will be described below with reference to relevant accompanying drawings and specific experiments. The specific content is as follows: 1. Experimental Environment Setup (1) Hardware configuration A mini PC equipped with an Intel 5300 Network Interface Controller (NIC) was used as the transmitter and receiver. The transmitter was equipped with one antenna and the receiver was equipped with three antennas with a spacing of 2.4 cm between adjacent antennas. The carrier frequency of the Wi-Fi channel was set to 5.32 GHz and the data packet rate was 100 Hz.

[0043] (2) Software configuration Channel status information was collected using the CSI tool, with the terminal operating system running Ubuntu 14.04.3. The data processing terminal was configured with an Intel Core i9-10900K processor and 64 GB of memory, on which MATLAB was used for data processing and positioning calculations.

[0044] After the devices are powered on and complete basic configuration, the host remotely controls the three devices via the SSH protocol. The transmitter is configured in inject mode with a packet interval of 10ms (100Hz packet rate; subsequent lower sampling rates can be achieved through downsampling), and signal transmission is controlled via command-line instructions. The two receivers are simultaneously configured in monitor mode to ensure simultaneous capture of the transmitter's channel status information. With this configuration, user movement within the feasible area of ​​the location can be accurately detected during the transmitter's effective packet transmission period.

[0045] (3) Scene setting Two scenarios were selected: an open space (5.6 m × 5.6 m) and an office. The positioning area was divided into grid cells of 0.1 m × 0.1 m. One transmitter (coordinate [0, 0]) and two receivers (coordinates [5.6, 0] and [0, 5.6] respectively) were deployed. The normal direction of the receiver antenna was consistent with the positive x-axis direction.

[0046] (4) Test object Five volunteers (three men and two women) sat statically in different positions within the designated area and used Neulog respiratory monitoring straps (model NUL236) to record their breathing status as a reference. The monitoring straps were connected to a PC via a USB-200 module.

[0047] 2. Implementation of the signal cleaning phase (1) Reference signal synthesis 1) Collect CSI data H(m, n, t) from 3 antennas over 30 subcarriers (n=1~30) for 30 seconds (t=1~3000, sampling interval 0.01 s), where antenna index m=1,2,3; 2) Define the expression for the reference signal as:

[0048] in For antenna weights, This represents the total phase shift. 3) Solving optimization problems using the traversal method To obtain the optimal weight ,in This is the average value of the reference signal at subcarrier n (characterizing the static signal energy). The variance of the reference signal at subcarrier n (characterizing the dynamic signal energy). 4) Substitute the optimal weights into the reference signal expression to synthesize the reference signal. The proportion of dynamic components in the synthesized reference signal is significantly reduced, and the static characteristics are stable.

[0049] (2) CFO and SFO clearance Calculate the ratio of each antenna's CSI data to the reference signal using the following formula:

[0050] This operation can eliminate the time-domain phase drift caused by CFO and the phase deviation between subcarriers caused by SFO, making the signal phase distribution closer to the ideal state.

[0051] 3. Implementation of the time-domain analysis phase (1) Respiratory signal filtering A bandpass filter was used to filter the respiratory signal. The passband frequency was set to 0.17 Hz-0.62 Hz. H'(m, n, t) was filtered to remove high-frequency noise, environmental noise and other irrelevant signals. The periodicity of the filtered signal was significantly enhanced and consistent with the respiratory cycle.

[0052] (2) Selection of the optimal projection angle Traverse the candidate projection angles φ∈[0, 2π), with a step size of 1°, and define the unit vector of the candidate projection angles as... The projection sequence of the CSI signal at each candidate projection angle is calculated using the following formula:

[0053] Where R(·) and I(·) represent the real and imaginary parts of the signal, respectively.

[0054] Finally, the Breath Noise Ratio (BNR) of each projection sequence is calculated. BNR is defined as the ratio of signal energy to total signal energy within the respiratory frequency range. The projection angle with the largest BNR is selected as the optimal projection angle.

[0055] (3) Projection angle correction For the original projection angle sequence Scaled by 4 times, we get Then calculate The phase is extracted, and the phase transition is smoothed using the periodicity of a complex function. The processed phase is divided by 4 to obtain the corrected projection angle sequence.

[0056] Where angle(·) represents the extraction of the phase of the complex function. After correction, the 90° jump phenomenon in the projection angle sequence disappears, and the continuity of the sequence is significantly improved.

[0057] 4. Implementation of frequency domain and spatial domain analysis phases (1) Multicarrier analysis (ToF extraction) This stage requires processing the corrected projection angle sequence. A linear fit is performed based on the subcarrier index n to obtain the slope k of the fitted line; the theoretical linear relationship between the projection angle and the subcarrier index is:

[0058] in For subcarrier spacing, The Time-of-Flight (ToF) of the dynamic signal (respiratory reflex signal) For the Time of Flight (ToF) of a static signal (direct signal), where C is a constant, the slope can be determined from this theoretical relationship. .

[0059] In this case, since the positions of the transmitter and receiver are known, the Time-of-Flight (ToF) of the static signal... Through formula Calculate (where For the transmitter location, (where c is the receiver position and c is the electromagnetic wave propagation speed). Using the slope k, the Time-of-Flight (ToF) of the dynamic signal can be calculated using the following formula: .

[0060] (2) Multi-antenna analysis (AoA extraction) Calculate the difference in the corrected projection angles between different antennas to obtain the total phase difference. Then, the average value of the total phase difference over 1 minute is measured to obtain the phase difference caused by the IPO. ,deduct Then, the phase difference corresponding to AoA is obtained. Based on the antenna array geometry, the theoretical formula for the phase difference corresponding to the signal arrival angle θ is:

[0061] in For carrier frequency (in this embodiment) ), d is the antenna spacing (d=0.024m in this embodiment), c is the electromagnetic wave propagation speed, and m is the antenna index. The formula for calculating AoA is obtained by transformation. Substituting the data, AoA can be calculated.

[0062] (3) IPO clearing Calculate the theoretical angle of arrival of the direct signal based on the positions of the transmitter and receiver. For example, when the angle is 0, it means that the antenna normal of receiver 1 points towards the transmitter. Substituting this into the phase difference theory formula, we can calculate the theoretical phase difference corresponding to the theoretical angle of arrival. The calculation formula is as follows: Substituting the data, we can calculate... Then, an IPO phase difference correction is performed, calculated using the following formula: This data is then subtracted from the phase data of each antenna to complete the precise removal of the IPO.

[0063] 5. Implementation of the location positioning phase A 5.6 m × 5.6 m positioning area was divided into 0.1 m × 0.1 m grids, resulting in 3136 candidate grid cells. The coordinates of each cell are... For each candidate unit, calculate the theoretical Time-of-Flight (ToF) and the theoretical AoA, where: The theoretical ToF calculation formula for receiver 1 is: ; The theoretical ToF calculation formula for receiver 2 is: ; The theoretical AoA calculation formula for receiver 1 is: ; The theoretical AoA calculation formula for receiver 2 is as follows: ; Compensation for the projection angle using theoretical Time-of-Flight (ToF) and AoA: For the nth subcarrier and the mth antenna, the compensation amount corresponding to ToF is... The compensation amount corresponding to AoA is Next, the variance of all projected angles after compensation is calculated. Select the sum of variances The smallest grid cell is used as the positioning result.

[0064] Repeat the above process and test 5 volunteers in different locations and orientations (-30°~30°). The median positioning error was 0.75 m in open space and 1.06 m in office, both of which meet the meter-level positioning requirements.

[0065] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and its improved concept, should be covered within the scope of protection of the present invention.

Claims

1. A user location positioning method based on human respiratory Wi-Fi reflection signals, characterized in that, Includes the following steps: S1. Signal Acquisition and Cleaning: The receiver acquires the raw signal containing the dynamic changes in channel state information caused by human respiration through multiple antennas; the raw signal is cleaned to eliminate carrier frequency offset and sampling frequency offset, and the cleaned channel state information ratio signal is obtained. S2. Time Domain Analysis: Bandpass filtering is performed on the cleaned channel state information ratio signal to extract the signal component corresponding to the breathing frequency; the optimal projection axis that maximizes the breathing periodicity is determined from the signal component, and the corresponding optimal projection angle sequence is calculated. S3. Frequency and Spatial Domain Analysis: Based on the optimal projection angle sequence, the time difference between the dynamic and static paths is extracted by analyzing the relationship between the optimal projection angle and the subcarrier index; the angle of arrival corresponding to the breathing signal is extracted by analyzing the relationship between the optimal projection angle and the receiving antenna index; combined with the known transceiver position information, the geometric information of the candidate reflection path corresponding to the breathing signal is calculated based on the time difference between the dynamic and static paths and the signal angle of arrival. S4. Positioning: Divide the positioning area into a grid; for each candidate grid position, calculate the flight time difference between the corresponding theoretical dynamic path and static path, as well as the theoretical signal angle of arrival, based on their geometric relationships; The measured optimal projection angle sequence is compensated by using the time difference between the theoretical dynamic path and the static path, as well as the theoretical signal angle of arrival. The overall variance of the compensated projection angle on all subcarriers and all antennas is calculated. The grid position that minimizes the overall variance is determined as the user's estimated position.

2. The method according to claim 1, characterized in that, The signal cleaning process described in S1 specifically includes: S11. Reference signal synthesis: The raw channel state information signals collected by multiple antennas of the receiver are weighted and summed to synthesize a reference signal. The weights are selected according to the criterion of maximizing the ratio of the signal average to the variance in order to suppress the breathing dynamic component in the reference signal. S12. Phase error elimination: Divide the original channel state information signals received by each antenna by the synthesized reference signal to obtain the channel state information ratio signal, so as to eliminate the common carrier frequency offset and sampling frequency offset phase error between different antennas.

3. The method according to claim 1, characterized in that, The determination of the optimal projection axis from the signal components as described in S2 is as follows: The filtered respiratory signal is mapped onto the complex plane. By calculating the variance or respiratory signal-to-noise ratio of the signal in different projection axis directions, the axis that maximizes the variance or respiratory signal-to-noise ratio is selected as the optimal projection axis. The angle between the optimal projection axis and the reference axis constitutes the optimal projection angle sequence.

4. The method according to claim 3, characterized in that, The S2 further includes using a complex smoothing algorithm to correct the optimal projection angle sequence in order to solve the phase jump problem caused by projection blur.

5. The method according to claim 1, characterized in that, As described in S3, the flight time difference between the dynamic path and the static path is extracted by analyzing the relationship between the optimal projection angle and the subcarrier index. Specifically, for the same moment, the optimal projection angle corresponding to different subcarrier indices is linearly fitted, and the slope of the fitted line is proportional to the flight time difference between the dynamic path and the static path.

6. The method according to claim 1, characterized in that, The method described in S3 for extracting the signal angle of arrival by analyzing the relationship between the optimal projection angle and the receiving antenna index specifically includes: S31. Obtain the optimal projection angle difference value of each receiving antenna relative to the reference antenna to form a spatial phase difference sequence; S32. Using the known geometric positional relationship of the transceiver, calculate and eliminate the influence of the initial phase offset caused by hardware on the spatial phase difference sequence; S33. Based on the geometric relationship between the spatial phase difference sequence after eliminating the initial phase offset and the antenna spacing, the signal arrival angle corresponding to the breathing signal is calculated.

7. The method according to claim 1, characterized in that, The method described in S4 for compensating the measured optimal projection angle sequence using the time difference between the theoretical dynamic path and the static path, as well as the theoretical signal angle of arrival, is as follows: For each subcarrier and each antenna, a theoretical phase compensation amount is calculated based on the time difference between the theoretical dynamic path and the static path, as well as the theoretical signal angle of arrival; the measured optimal projection angle is subtracted from the corresponding theoretical phase compensation amount to obtain the compensated projection angle.

Images