Indoor positioning and identity recognition method based on channel state information

Abstract

The invention relates to an indoor positioning and identity recognition method based on channel state information. A to-be-detected channel state information signal is obtained from an array antenna of a receiving device; correcting the phase of the obtained channel state information signal; performing static and dynamic signal separation on the channel state information after phase correction; the motion speed v, the angle of arrival AOA and the time of flight TOF of the moving target are obtained through a maximum likelihood estimation algorithm according to the obtained signals of the dynamic path; weighting the current TOF value according to the speed at the previous moment to obtain a more accurate TOF value; the obtained more accurate TOF value and the obtained AOA value are positioned through particle filtering, the respiratory rate, heart rate detection, relative body type energy and relative gait cycle of the moving target are obtained and substituted into a model trained in advance, and the identity of the moving target is obtained according to the maximum probability. The method has the advantages of being universal in equipment, convenient to erect, simple in algorithm, capable of being used for real-time tracking and positioning and convenient to popularize on a large scale.

Description

Technical Field

[0001] The present invention relates to the field of wireless network and identification technology, in particular to an indoor positioning and identity recognition method based on channel state information. Background Art

[0002] With the development and maturity of 5G wireless communication technology and the continuous evolution of 6G technology, indoor positioning technology has developed rapidly. Indoor location perception technology has been widely used in special population positioning, large venue management, Internet of Things, personal location services and other fields. Indoor location information is particularly important in emergency scenarios, such as fire rescue, emergency evacuation, earthquake relief, etc. However, due to the complex indoor environment, the multipath effect is obvious, and the fingerprint positioning algorithm based on deep learning is difficult to adapt to environmental changes; in addition, the common indoor positioning algorithm based on Wi-Fi signals requires multiple groups of transceiver links, which is less applicable to ordinary families; in addition, the traditional parameter-based positioning algorithm has low resolution, complex algorithm, and it is difficult to meet the needs of real-time positioning. Identity recognition plays an important role in home-based elderly care, anti-theft, security and other aspects. It is still a major challenge at home and abroad to realize universal, real-time, efficient, and high-precision positioning algorithms and identify the identity of moving targets at the same time. In response to the above problems, this system proposes a low-cost, high-precision location perception, real-time trajectory tracking, and identity recognition system suitable for the public using the channel state information (CSI) of Wi-Fi signals. Summary of the invention

[0003] The purpose of the present invention is to provide a method for indoor positioning and identity recognition based on channel state information, which can realize universal equipment, convenient installation, simple algorithm, and can be used for real-time tracking and positioning.

[0004] To achieve the above object, the present invention provides the following technical solution: a method for indoor positioning and identity recognition based on channel state information, the method comprising the following steps:

[0005] Step S1, start the transmitting device and the receiving device, the network card and the computer to work, and obtain the channel state information signal to be detected from the array antenna of the receiving device;

[0006] Step S2, phase correcting the obtained channel state information signal; performing static and dynamic signal separation on the phase-corrected channel state information according to spatial spectrum estimation theory;

[0007] Step S3, the obtained dynamic path signal is used to obtain the moving target's moving speed v, arrival angle AOA, and flight time TOF using the maximum likelihood estimation algorithm; then the current TOF value is weighted according to the speed at the previous moment to obtain a more accurate TOF value;

[0008] Step S4, using particle filtering to locate the more accurate TOF value and the obtained arrival angle AOA value, obtaining the breathing rate and heart rate detection of the moving target from the dynamic signal, obtaining the relative body energy of the moving target, obtaining the relative gait cycle of the moving target, substituting the obtained breathing rate, heart rate detection, relative body energy, and relative gait cycle of the moving target into the pre-trained model, and obtaining the identity of the moving target according to the maximum probability. The pre-trained model can use a known neural network model (including but not limited to CNN, RNN, LSTM, self-attention mechanism, etc.).

[0009] Furthermore, the channel state information signal to be detected obtained from the array antenna of the receiving device in step S1 is further: the channel state information CSI of the received wireless signal Wi-Fi signal reflects the information of static and dynamic objects in the environment, which is expressed as:

[0010]

[0011] Here L represents the total number of dynamic and static paths in the environment. l (i,k) represents the CSI signal of the i-th data packet of the k-th subcarrier on the l-th path. n(i,k) represents the noise in the environment. l (i,k) represents the amplitude of the lth path, τ l represents the flight time of the lth path. f represents the carrier frequency.

[0012] Furthermore, the static and dynamic signal separation of the phase-corrected channel state information according to the spatial spectrum estimation theory in step S2 is further as follows:

[0013] According to the spatial spectrum estimation theory, the signal feature space U can be divided into the signal space U S and noise space U N :

[0014] R=UΣU H =U s Σ s U s H +U N Σ N U N H ,

[0015] The signal space U S Divided into static signal space U L and the dynamic signal space U D :

[0016] R=U D Σ D UD H +U L Σ L U L H +U N Σ N U N H ;

[0017] According to the decomposition of the eigenvector space, the eigenvalue Σ of the total signal R space is composed of the signal space eigenvalue Σ S and the noise space eigenvalue Σ N Composition (where Σ contains the eigenvalue λ 1 , 2 ... M ,Σ S Contains eigenvalue λ 1 , 2 ... N ,Σ N Contains eigenvalue λ N+1 , N+2 ... M );

[0018]

[0019] h l The static path S contained in (i,k) L It can be expressed as:

[0020]

[0021] Because the static paths differ only by a complex constant, they are coherent and merge into one path. The eigenvalue of the signal covariance represents the power of the signals of different paths. Therefore, the signal with the largest eigenvalue in the signal space is the static signal, and the remaining N-1 paths are dynamic signals. 静 represents the characteristic value of the static path signal, λ 动 represents the characteristic value of the dynamic path signal, that is,

[0022]

[0023] Furthermore, the step S3 is further as follows: the CSI signal of the dynamic path to is divided into a plurality of short-time windows, and the average speed v, the angle of arrival AOA, and the flight time TOF in the sliding window are regarded as constants, t=1, 2, 3, ..., T represents the number of sliding windows; after segmentation, the speed v of the moving target is used as the i,k,m Approximately meets the following conditions:

[0024] v i,k,m =v t +ε.

[0025] where v represents the average speed of window t t , ε is the estimated noise, ε~N(0,σ 1 ) has a mean of 0 and a variance of σ 1 Normal distribution of

[0026] By (v i,k,m |v t ) formula to obtain the speed v of the ith data packet of the kth subcarrier of the mth antenna in the short time window i,k,m ; then the probability density function of its velocity can be expressed as:

[0027]

[0028] I represents the total number of data packets, K represents the total number of subcarriers contained in each antenna, M represents the total number of antennas, and the joint probability distribution function of all speed observations is:

[0029]

[0030] The maximum likelihood estimation result of the radial velocity relative to the ellipse with the transceiver as the focus is to first take the logarithm of the joint probability distribution function and then maximize the joint probability distribution function of the velocity observations:

[0031]

[0032] The distribution function In(p(v|v t )) t Taking the derivative and setting it to 0, we can get the maximum likelihood estimate of the radial velocity:

[0033]

[0034] For the arrival angle AOA, flight time TOF value tx q-1 The same method is used to obtain;

[0035] According to the speed of the previous moment, the current TOF value tx q-1 Weighted to get a more accurate TOF value tx q :

[0036]

[0037] where v mean represents the average speed at the previous moment, c represents the speed of light, and represents the time interval between the current moment and the previous moment.

[0038] Furthermore, the respiratory rate and heart rate detection of the moving target obtained in step S4 is further as follows: the signal after singular point filtering is decomposed by wavelet transform, Haar wavelet is selected, threshold processing is performed on the decomposed parameters, a suitable threshold is selected for analysis, and wavelet inverse transform is used for wavelet reconstruction to obtain the motion signal envelope w(t);

[0039] Subtract the motion signal envelope from the signal after singular point filtering to obtain the respiration and heart rate signals;

[0040] r(t)+h(t)=s(t)-w(t)

[0041] S(t) is the signal after row singular point filtering, w(t) is the motion signal envelope, r(t) is the breathing signal, and h(t) is the heart rate signal;

[0042] The acquired respiration and heart rate signals are subjected to Butterworth bandpass filtering, with a passband frequency of 0.15-0.5 Hz to filter out the respiration signal, and a passband frequency of 0.8-2.5 Hz to filter out the heart rate signal; a discrete Fourier transform is performed, and the frequency corresponding to the peak point detected in the spectrum is the respiration rate and heart rate.

[0043] Furthermore, the relative body energy of the moving target obtained in step S4 is further:

[0044] Each antenna can obtain 30 subcarriers, and the three receiving antennas receive a total of 90 subcarriers. The average energy of the 90 subcarriers is calculated as follows:

[0045]

[0046] S(t) is the signal after row singular point filtering, S(jw) is the spectrum of S(t) signal, and E is the energy of S(t) signal;

[0047] Because distance has a great influence on energy, in order to more accurately reflect the size of the moving target, the distance d(t) between the target and the receiving antenna is calculated according to the located position and the target energy is weighted:

[0048] E(t) 相对 =E / d(t)

[0049] Furthermore, the relative gait period of the moving target obtained in step S4 is further:

[0050] The obtained motion signal envelope w(t) is subjected to discrete Fourier transform, and the frequency corresponding to the peak point detected in the spectrum is the gait period B(t). Because distance has a great influence on the gait period, in order to more accurately reflect the size of the moving target, the distance d(t) between the target and the receiving antenna is calculated based on the obtained position and the target gait period is weighted:

[0051] B(t) 相对 =B(t) / d(t).

[0052] Furthermore, in step S4, the more accurate TOF value and the obtained arrival angle AOA value are used to realize positioning by particle filtering, and further: according to the angle θ reaching the array antenna after being reflected by the target 1 , the unique position can be determined within the detection area. If the target position is (x, y), the transmitting antenna F 1 and receiving antenna F 2 The positions of (x 1 ,y 1 ) and (x 2 ,y 2 ), the target coordinates can be obtained using the following equation:

[0053]

[0054] Beneficial effects of the present invention: The present invention can realize accurate real-time positioning and identify the identity of the target through a set of sending and receiving devices, and can accurately separate various signals of the moving target from signals containing complex noise and multipath. The positioning process uses the maximum likelihood estimation algorithm to calculate the moving speed of the moving target, the arrival angle relative to the receiving antenna array, and the flight time of the signal from the sending device to the receiving device after being reflected by the moving target. Finally, an adaptive particle filter is used to achieve real-time positioning. The identity recognition process mainly collects information such as the frequency and amplitude of the moving target's breathing, arms and legs, etc. to achieve identity recognition. The system has the advantages of being easy to set up and easy to promote on a large scale. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a schematic diagram of the AOA and TOF positioning algorithm implemented by the present invention using the WiFi channel state information;

[0056] Figure 2 Flowchart of indoor positioning and identification system using WiFi channel status information; DETAILED DESCRIPTION

[0057] The present invention will be further described below in conjunction with the accompanying drawings.

[0058] See also Figure 1 Figure 2As shown, a method for indoor positioning and identity recognition based on channel state information of the present invention comprises the following steps:

[0059] Step S1: Start the transmitting device, receiving device, network card and computer to work, and obtain the channel state information signal to be detected from the array antenna of the receiving device.

[0060] Step S2, phase correcting the obtained channel state information signal; performing static and dynamic signal separation on the phase-corrected channel state information according to spatial spectrum estimation theory;

[0061] Step S3, the obtained dynamic path signal is used to obtain the moving target's moving speed v, arrival angle AOA, and flight time TOF using the maximum likelihood estimation algorithm; then the current TOF value is weighted according to the speed at the previous moment to obtain a more accurate TOF value;

[0062] Step S4, using particle filtering to achieve positioning with the more accurate TOF value and the obtained arrival angle AOA value, obtaining the breathing rate and heart rate detection of the moving target from the dynamic signal, obtaining the relative body energy of the moving target, obtaining the relative gait cycle of the moving target, substituting the obtained breathing rate, heart rate detection, relative body energy, and relative gait cycle of the moving target into the pre-trained model, and obtaining the identity of the moving target according to the maximum probability.

[0063] The present invention will be further described below in conjunction with a specific embodiment:

[0064] A method for indoor positioning and identification based on channel state information,

[0065] Step S1, start working of the transmitting device and the receiving device, the network card and the computer, and obtain the channel state information signal to be detected from the array antenna of the receiving device; the wireless signal includes but is not limited to 2.4GHZ and 5.8GHZ WiFi signals, and the amplitude and phase information of the radio signal will be affected by the moving target during the propagation process from the WiFi signal transmitter to the receiver. The channel state information collected by the receiving end contains all the information of the moving target in a certain space, including the moving speed, position, height, body shape information, etc. of the moving target.

[0066] The channel state information signal to be detected is obtained from the array antenna of the receiving device in step S1 further: the channel state information CSI of the received wireless signal Wi-Fi signal reflects the information of static and dynamic objects (i.e., paths) in the environment, which is expressed as:

[0067]

[0068] Here L represents the total number of dynamic and static paths in the environment.l (i,k) represents the CSI signal of the i-th data packet of the k-th subcarrier on the l-th path. n(i,k) represents the noise in the environment. l (i,k) represents the amplitude of the lth path, τ l represents the flight time of the lth path. f represents the carrier frequency.

[0069] Step S2, phase correction of the obtained channel state information signal; the phase correction of the obtained channel state information signal is due to the fact that there is no strict clock synchronization in the reception process of the CSI signal, and there will be errors between the measured phase and the actual phase. According to the spatial spectrum estimation theory, the static and dynamic signals of the phase-corrected channel state information are separated; this can remove the signals formed by reflections from the ground, furniture, etc., which are useless for subsequent processing and will also cause interference;

[0070] The step S2 of performing static and dynamic signal separation on the phase-corrected channel state information according to the spatial spectrum estimation theory is further as follows:

[0071] According to the spatial spectrum estimation theory, the signal feature space U can be divided into the signal space U S and noise space U N :

[0072] R=UΣU H =U s Σ s U s H +U N Σ N U N H ,

[0073] The signal space U S Divided into static signal space U L and the dynamic signal space U D :

[0074] R=U D Σ D U D H +U L Σ L U L H +U N Σ N U N H ;

[0075] According to the decomposition of the eigenvector space, the eigenvalue Σ of the total signal R space is composed of the signal space eigenvalue Σ S and the noise space eigenvalue Σ NComposition (where Σ contains the eigenvalue λ 1 , 2 ... M ,Σ S Contains eigenvalue λ 1 , 2 ... N ,Σ N Contains eigenvalue λ N+1 , N+2 ... M );

[0076]

[0077] h l The static path S contained in (i,k) L It can be expressed as:

[0078]

[0079] Because the static paths differ only by a complex constant, they are coherent and merge into one path. The eigenvalue of the signal covariance represents the power of the signals of different paths. Therefore, the signal with the largest eigenvalue in the signal space is the static signal, and the remaining N-1 paths are dynamic signals. 静 represents the characteristic value of the static path signal, λ 动 represents the characteristic value of the dynamic path signal, that is,

[0080]

[0081] Step S3, the obtained dynamic path signal, respectively, uses the maximum likelihood estimation algorithm to obtain the moving target's moving speed v, arrival angle AOA, and flight time TOF; then the current TOF value is weighted according to the speed at the previous moment to obtain a more accurate TOF value; the method of the present invention is essentially different from Widar2.0. Widar2.0 mainly constructs an optimization function, traverses the amplitude, AOA, TOF, and Doppler speed, and finds the parameters corresponding to the minimum square error between the estimated CSI signal and the measured CSI signal. This patent is mainly based on the assumption that the estimated deviation (noise) of the three parameters of the moving target's moving speed v, arrival angle AOA, and flight time TOF meets the normal distribution with a mean of 0.

[0082] The step S3 is further as follows: the CSI signal of the dynamic path to the target is divided into multiple short-term windows, and the average speed v, arrival angle AOA, and flight time TOF in the sliding window are regarded as constants, t=1, 2, 3, ..., T represents the number of sliding windows; after segmentation, the speed v of the moving target is used as the i,k,m Approximately meets the following conditions:

[0083] vi,k,m =v t +ε.

[0084] where v represents the average speed of window t t , ε is the estimated noise, ε~N(0,σ 1 ) has a mean of 0 and a variance of σ 1 Normal distribution of

[0085] By (v i,k,m |v t ) formula to obtain the speed v of the ith data packet of the kth subcarrier of the mth antenna in the short time window i,k,m ; then the probability density function of its velocity can be expressed as:

[0086]

[0087] I represents the total number of data packets, K represents the total number of subcarriers contained in each antenna, M represents the total number of antennas, and the joint probability distribution function of all speed observations is:

[0088]

[0089] The maximum likelihood estimation result of the radial velocity relative to the ellipse with the transceiver as the focus is to first take the logarithm of the joint probability distribution function and then maximize the joint probability distribution function of the velocity observations:

[0090]

[0091] The distribution function In(p(v|v t )) t Taking the derivative and setting it to 0, we can get the maximum likelihood estimate of the radial velocity:

[0092]

[0093] For the arrival angle AOA, flight time TOF value tx q-1 The same method is used to obtain;

[0094] According to the speed of the previous moment, the current TOF value tx q-1 Weighted to get a more accurate TOF value tx q :

[0095]

[0096] where v mean represents the average speed at the previous moment, c represents the speed of light, and represents the time interval between the current moment and the previous moment.

[0097] Step S4, using particle filtering to achieve positioning with the more accurate TOF value and the obtained arrival angle AOA value, obtaining the breathing rate and heart rate detection of the moving target from the dynamic signal, obtaining the relative body energy of the moving target, obtaining the relative gait cycle of the moving target, substituting the obtained breathing rate, heart rate detection, relative body energy, and relative gait cycle of the moving target into the pre-trained model, and obtaining the identity of the moving target according to the maximum probability.

[0098] If the AOA and TOF values ​​are obtained, the position is unique within the detection area, as follows Figure 1 Show.

[0099] The respiratory rate and heart rate detection of the moving target obtained in step S4 is further performed by: decomposing the signal after singular point filtering by wavelet transform, selecting Haar wavelet, performing threshold processing on the decomposed parameters, selecting a suitable threshold for analysis, and performing wavelet reconstruction by inverse wavelet transform to obtain the motion signal envelope w(t);

[0100] Subtract the motion signal envelope from the signal after singular point filtering to obtain the respiration and heart rate signals;

[0101] r(t)+h(t)=s(t)-w(t)

[0102] S(t) is the signal after row singular point filtering, w(t) is the motion signal envelope, r(t) is the breathing signal, and h(t) is the heart rate signal;

[0103] The acquired respiration and heart rate signals are subjected to Butterworth bandpass filtering, with a passband frequency of 0.15-0.5 Hz to filter out the respiration signal, and a passband frequency of 0.8-2.5 Hz to filter out the heart rate signal; a discrete Fourier transform is performed, and the frequency corresponding to the peak point detected in the spectrum is the respiration rate and heart rate.

[0104] The relative body energy of the moving target obtained in step S4 is further:

[0105] Each antenna can obtain 30 subcarriers, and the three receiving antennas receive a total of 90 subcarriers. The average energy of the 90 subcarriers is calculated as follows:

[0106]

[0107] S(t) is the signal after row singular point filtering, S(jw) is the spectrum of S(t) signal, and E is the energy of S(t) signal;

[0108] Because distance has a great influence on energy, in order to more accurately reflect the size of the moving target, the distance d(t) between the target and the receiving antenna is calculated according to the located position and the target energy is weighted:

[0109] E(t) 相对 =E / d(t)

[0110] The relative gait period of the moving target obtained in step S4 is further:

[0111] The obtained motion signal envelope w(t) is subjected to discrete Fourier transform, and the frequency corresponding to the peak point detected in the spectrum is the gait period B(t). Because distance has a great influence on the gait period, in order to more accurately reflect the size of the moving target, the distance d(t) between the target and the receiving antenna is calculated based on the obtained position and the target gait period is weighted:

[0112] B(t) 相对 =B(t) / d(t).

[0113] In step S4, the more accurate TOF value and the obtained arrival angle AOA value are further positioned by using particle filtering: according to the angle θ reaching the array antenna after being reflected by the target 1 , the unique position can be determined within the detection area. If the target position is (x, y), the transmitting antenna F 1 and receiving antenna F 2 The positions of (x 1 ,y 1 ) and (x 2 ,y 2 ), the target coordinates can be obtained using the following equation:

[0114]

[0115] The present invention has the following working principle: In short, the present invention can realize accurate real-time positioning and identify the identity of the target through a set of sending and receiving devices, and can accurately separate various signals of the moving target from signals containing complex noise and multipath. The positioning process uses the maximum likelihood estimation algorithm to calculate the moving speed of the moving target, the arrival angle relative to the receiving antenna array, and the flight time of the signal from the sending device to the receiving device after being reflected by the moving target. Finally, an adaptive particle filter is used to realize real-time positioning. The identity recognition process mainly collects information such as the frequency and amplitude of the moving target's breathing, arms and legs, etc. to realize identity recognition. The system has the advantages of being easy to set up and easy to promote on a large scale.

[0116] The above description is only a preferred embodiment of the present invention and should not be construed as limiting the present application. All equivalent changes and modifications made within the scope of the patent application of the present invention should fall within the scope of the present invention.

Claims

1. A method for indoor positioning and identification based on channel state information, characterized in that: The method comprises the following steps: Step S1, start the transmitting device and the receiving device, the network card and the computer to work, and obtain the channel state information signal to be detected from the array antenna of the receiving device; Step S2, phase correcting the obtained channel state information signal; performing static and dynamic signal separation on the phase-corrected channel state information according to spatial spectrum estimation theory; Step S3, the obtained dynamic path signal is used to obtain the moving target's moving speed v, arrival angle AOA, and flight time TOF using the maximum likelihood estimation algorithm; then the current TOF value is weighted according to the speed at the previous moment to obtain a more accurate TOF value; Step S4, using particle filtering to achieve positioning with the more accurate TOF value and the obtained arrival angle AOA value, obtaining the breathing rate and heart rate detection of the moving target from the dynamic signal, obtaining the relative body energy of the moving target, obtaining the relative gait cycle of the moving target, substituting the obtained breathing rate, heart rate detection, relative body energy, and relative gait cycle of the moving target into the pre-trained model, and obtaining the identity of the moving target according to the maximum probability.

2. The method for indoor positioning and identification based on channel state information according to claim 1, characterized in that: The step S1 of obtaining the channel state information signal to be detected from the array antenna of the receiving device is further as follows: the channel state information CSI of the received wireless signal Wi-Fi signal reflects the information of static and dynamic objects in the environment, which is expressed as: Here L represents the total number of dynamic and static paths in the environment. l (i,k) represents the CSI signal of the i-th data packet of the k-th subcarrier on the l-th path. n(i,k) represents the noise in the environment. l (i,k) represents the amplitude of the lth path, τ l represents the flight time of the lth path. f represents the carrier frequency.

3. The method for indoor positioning and identification based on channel state information according to claim 1, characterized in that: The step S2 of performing static and dynamic signal separation on the phase-corrected channel state information according to the spatial spectrum estimation theory is further as follows: According to the spatial spectrum estimation theory, the signal feature space U can be divided into the signal space U S and noise space U N : R=UΣU H =U s Σ s IN s H +U N Σ N IN N H , The signal space U S Divided into static signal space U L and the dynamic signal space U D : R=U D Σ D IN D H +U L Σ L IN L H +U N Σ N IN N H ; According to the decomposition of the eigenvector space, the eigenvalue Σ of the total signal R space is composed of the signal space eigenvalue Σ S and the noise space eigenvalue Σ N Composition (where Σ contains eigenvalues ​​λ1, λ2, ... M ,Σ S Contains eigenvalues ​​λ1, λ2, ...λ N ,Σ N Contains eigenvalue λ N+1 , N+2 ... M ); h l The static path S contained in (i,k) L It can be expressed as: Because the static paths differ only by a complex constant, they are coherent and merge into one path. The eigenvalue of the signal covariance represents the power of the signals of different paths. Therefore, the signal with the largest eigenvalue in the signal space is the static signal, and the remaining N-1 paths are dynamic signals. 静 represents the characteristic value of the static path signal, λ 动 represents the characteristic value of the dynamic path signal, that is, 4. The method for indoor positioning and identification based on channel state information according to claim 1, characterized in that: The step S3 is further as follows: the CSI signal of the dynamic path to the target is divided into multiple short-term windows, and the average speed v, arrival angle AOA, and flight time TOF in the sliding window are regarded as constants, t=1, 2, 3, ..., T represents the number of sliding windows; after segmentation, the speed v of the moving target is used as the i,k,m Approximately meets the following conditions: v i,k,m =v t +ε。 where v represents the average speed of window t t , ε is the estimated noise, ε~N(0,σ1) obeys the normal distribution with mean 0 and variance σ1; By (v i,k,m |v t ) formula to obtain the speed v of the ith data packet of the kth subcarrier of the mth antenna in the short time window i,k,m ; then the probability density function of its velocity can be expressed as: I represents the total number of data packets, K represents the total number of subcarriers contained in each antenna, M represents the total number of antennas, and the joint probability distribution function of all speed observations is: The maximum likelihood estimation result of the radial velocity relative to the ellipse with the transceiver as the focus is to first take the logarithm of the joint probability distribution function and then maximize the joint probability distribution function of the velocity observations: The distribution function In(p(v|v t )) t Taking the derivative and setting it to 0, we can get the maximum likelihood estimate of the radial velocity: For the arrival angle AOA, flight time TOF value tx q-1 The same method is used to obtain; According to the speed of the previous moment, the current TOF value tx q-1 Weighted to get a more accurate TOF value tx q : where v mean represents the average speed at the previous moment, c represents the speed of light, and represents the time interval between the current moment and the previous moment.

5. The method for indoor positioning and identification based on channel state information according to claim 1, characterized in that: The respiratory rate and heart rate detection of the moving target obtained in step S4 is further performed by: decomposing the signal after singular point filtering by wavelet transform, selecting Haar wavelet, performing threshold processing on the decomposed parameters, selecting a suitable threshold for analysis, and performing wavelet reconstruction by inverse wavelet transform to obtain the motion signal envelope w(t); Subtract the motion signal envelope from the signal after singular point filtering to obtain the respiration and heart rate signals; r(t)+h(t)=s(t)-w(t) S(t) is the signal after row singular point filtering, w(t) is the motion signal envelope, r(t) is the breathing signal, and h(t) is the heart rate signal; The acquired respiration and heart rate signals are subjected to Butterworth bandpass filtering, with a passband frequency of 0.15-0.5 Hz to filter out the respiration signal, and a passband frequency of 0.8-2.5 Hz to filter out the heart rate signal; a discrete Fourier transform is performed, and the frequency corresponding to the peak point detected in the spectrum is the respiration rate and heart rate.

6. The method for indoor positioning and identification based on channel state information according to claim 1, characterized in that: The relative body energy of the moving target obtained in step S4 is further: Each antenna can obtain 30 subcarriers, and the three receiving antennas receive a total of 90 subcarriers. The average energy of the 90 subcarriers is calculated as follows: S(t) is the signal after row singular point filtering, S(jw) is the spectrum of S(t) signal, and E is the energy of S(t) signal; Because distance has a great influence on energy, in order to more accurately reflect the size of the moving target, the distance d(t) between the target and the receiving antenna is calculated according to the located position and the target energy is weighted: E(t) 相对 =E / d(t)。 7. The method for indoor positioning and identification based on channel state information according to claim 1, characterized in that: The relative gait period of the moving target obtained in step S4 is further: The obtained motion signal envelope w(t) is subjected to discrete Fourier transform, and the frequency corresponding to the peak point detected in the spectrum is the gait period B(t). Because distance has a great influence on the gait period, in order to more accurately reflect the size of the moving target, the distance d(t) between the target and the receiving antenna is calculated based on the obtained position and the target gait period is weighted: B(t) 相对 =B(t) / d(t)。 8. The method for indoor positioning and identification based on channel state information according to claim 1, characterized in that : In step S4, the more accurate TOF value and the obtained arrival angle AOA value are further positioned using particle filtering: According to the angle θ1 reaching the array antenna after being reflected by the target, a unique position can be determined within the detection area. If the target position is (x, y), the positions of the transmitting antenna F1 and the receiving antenna F2 are (x1, y1) and (x2, y2) respectively, the target coordinates can be obtained using the following equation: Where c represents the speed of light.

Images