Indoor fingerprint positioning method based on terahertz channel characteristics

Abstract

The invention relates to an indoor fingerprint positioning method based on terahertz channel characteristics, and belongs to the technical field of wireless positioning. The method comprises the following steps: firstly, collecting terahertz channel response information at different positions, extracting channel characteristic parameters with stability and distinction degree, and constructing an indoor positioning fingerprint database; using a DBSCAN algorithm to form a clustering fingerprint database; terahertz channel information of a to-be-positioned terminal is obtained in real time, channel features of the same type are extracted and matched with samples in a fingerprint database, and therefore position estimation is achieved. According to the method, the high resolution, the multipath characteristic and the position sensitivity of the terahertz channel in the indoor propagation process are fully utilized, the precision and the robustness of a positioning system are improved, and the method is suitable for high-precision positioning requirements in various complex indoor environments.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless positioning, and relates to an indoor fingerprint positioning method based on terahertz channel characteristics. Background Art

[0002] With the rapid development of the Internet of Things, smart buildings, and mobile services, indoor positioning technology, as one of the important foundations to support the construction of smart cities, has received extensive attention. Currently, the mainstream indoor positioning methods mainly include technologies based on received signal strength (RSSI), time of arrival (TOA), angle of arrival (AOA), and wireless fingerprints. However, limited by problems such as narrow bandwidth, low resolution, and poor multipath resistance of existing wireless communication bands (such as Wi-Fi, Bluetooth, UWB, etc.), traditional positioning methods are difficult to achieve high-precision positioning in complex indoor environments. In recent years, terahertz communication technology, due to its extremely high frequency (0.1–10 THz), ultra-wide bandwidth, and strong penetration and reflection capabilities, is considered an important development direction for the next generation of wireless communication and sensing integration. The terahertz channel exhibits more significant spatial resolution characteristics and channel fingerprint uniqueness during propagation, with higher potential for positioning accuracy, especially suitable for complex indoor environments with high density and multiple reflections. The fingerprint positioning method based on terahertz channel characteristics can effectively utilize the time-frequency characteristics of the channel response to construct a finer and more stable location fingerprint database, breaking through the accuracy bottleneck of traditional positioning methods and becoming one of the current research hotspots.

[0003] However, the existing indoor positioning research based on terahertz channels is still in its infancy, and there is still a lack of methods for in-depth modeling of channel characteristics and effective fingerprint feature extraction.

[0004] Therefore, there is an urgent need to propose an indoor fingerprint positioning method that makes full use of terahertz channel information, has high stability and positioning accuracy, to meet the requirements for high-precision and robust indoor positioning in future intelligent environments. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide an indoor fingerprint positioning method based on terahertz channel characteristics, which makes full use of the rich channel characteristics of the terahertz band to construct a finer and more stable fingerprint database, and solves the problem of low positioning accuracy existing in current fingerprint positioning methods due to the fluctuation of RSSI.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] An indoor fingerprint positioning method based on terahertz channel characteristics, the method comprising the following steps:

[0008] S1. Divide the area to be located into grids, establish a coordinate system, and record the coordinate information of each grid point position;

[0009] S2. Obtain the terahertz channel data at each point through a terahertz channel measurement device and save the data;

[0010] S3. Process the terahertz channel data set using physical and statistical methods to obtain terahertz channel characteristics; combine the terahertz channel characteristics with coordinate information to construct a multi-dimensional fingerprint database;

[0011] S4. Use the density-based spatial clustering of applications with noise (DBSCAN) algorithm to cluster the multi-dimensional fingerprint database to obtain a clustered fingerprint database;

[0012] S5. Obtain the fingerprint features of the point to be located, use the weighted K-nearest neighbor matching algorithm to match the fingerprint features of the point to be located with the fingerprint features in the clustered fingerprint database, and output the two-dimensional coordinate values of the matching result to complete the positioning.

[0013] Further, in step S1, the process of establishing the coordinate system is as follows: The coordinate origin is located at the corner position in the southeast direction of the room, and the interval between coordinates is 0.8 m.

[0014] Further, in step S2, the terahertz channel measurement device measures using the grid points divided by the frequency-domain measurement method based on a vector network analyzer; among them, the terahertz channel measurement device includes a transmitter Tx, a receiver Rx, two radio frequency front-ends RF, a signal generator, a two-way power divider, and a vector network analyzer VNA.

[0015] Further, the process by which the terahertz channel measurement device obtains the terahertz channel data and saves the data includes:

[0016] S21. Place the transmitter Tx at the coordinate origin of the area to be located, and randomly place the two receivers Rx at the divided grid points and the point to be located to obtain the terahertz channel data set of the grid points and the terahertz channel data set of the point to be located;

[0017] S22. Amplify the local oscillator signal generated by the signal generator of the transmitter Tx to a preset frequency through a frequency multiplier, mix it with the intermediate frequency signal generated by the vector network analyzer VNA to a preset frequency band, and then the terahertz signal is transmitted by the horn antenna of the transmitter and received by the horn antenna of the receiver; at the same time, the signal generator of the receiver Rx generates a local oscillator signal of 13.104 GHz, and down-converts the terahertz signal to an intermediate frequency signal of a preset frequency; send the intermediate frequency signal back to the vector network analyzer VNA to obtain the complex amplitude ratio parameter, where the complex amplitude ratio parameter refers to the complex amplitude ratio when the signal is incident from the port and when it is output from the port, and is used to describe the forward transmission characteristics of the device under test;

[0018] S23. Install the transmitter Tx and the receiver Rx on the multi-dimensional turntable in a specific manner. Specifically, complete the installation with the half-power beam widths of the directional horn antennas of the transmitter Tx and the receiver Rx as the installation adjustment parameters. At the start of the measurement, the transmitter Tx remains stationary, and the receiver Rx rotates within the preset ranges in the H-plane and the E-plane and receives multipath components at a preset step size to complete the measurement.

[0019] Furthermore, during the measurement process using the terahertz channel measurement device, its parameter configuration includes: the center frequency of the terahertz measurement system, the local oscillator signal frequency of the signal generator, the frequency multiplication factor of the frequency multiplier, the intermediate frequency signal frequency generated by the VNA, the start frequency, the cut-off frequency, and the bandwidth range, the intermediate frequency bandwidth, the number of sweep points, the maximum power of IF IN, the transmitter antenna gain, the receiver antenna gain, the transmitter half-power beam width, the receiver half-power beam width, the delay resolution, the maximum remaining delay of the multipath components, the receiver azimuth range and elevation angle range, the sweep interval, and the maximum power of LO IN.

[0020] Furthermore, in step S3, the obtained terahertz channel characteristic parameters include large-scale parameters and small-scale parameters based on the environment. Among them, the large-scale parameters include the maximum power P m and the path loss PL at each grid point; the small-scale parameters include the delay spread RMSDS and the angle spread RMSAS at each grid point.

[0021] Furthermore, the calculation method of the maximum power P m at each grid point of the large-scale parameters is as follows:

[0022] PDAP(τ, θ RX , φ RX ) = |h channel (τ, θ RX , φ RX )| 2

[0023]

[0024] P m = 10lg(max(PDP(τ)))

[0025] where τ represents the delay, and PDP(·) represents the basic channel parameter derived from the channel impulse response used to characterize the power distribution characteristics of the signal in the time domain; θ RX is the elevation angle of the horn antenna, φ RX is the azimuth angle of the horn antenna; h channel (·) represents the impulse response of the channel;

[0026] The calculation method of the path loss PL at each grid point of the large-scale parameters is as follows:

[0027]

[0028] PL = G TX +G RX +L c +PL FS

[0029] Among them, PL Fs represents the path loss in free space, d is the distance between the transceiver, f is the frequency, c is the speed of light, G TX and G RX are the transmitter gain and the receiver gain respectively, and L c is the loss of the cable;

[0030] The root mean square delay spread (RMSDS) of each grid point of the small-scale parameter is represented by the second central moment of the PDP(·), and the calculation method is as follows:

[0031]

[0032] The calculation method of the root mean square angular spread (RMSAS) of each grid point of the small-scale parameter is:

[0033]

[0034] Among them, τ l represents the delay of the l-th multipath, and ψ l represents the direction of arrival (DOA) of the l-th multipath, which includes the azimuth of arrival (AOA) and the elevation of arrival (EOA).

[0035] Furthermore, in step S4, after clustering the database into a clustering fingerprint database by the DBSCAN clustering algorithm, the database is divided into multiple small regions, and the process is as follows:

[0036] By adjusting the neighborhood radius ε and the number of points Minpts included in a cluster to find the optimal parameters, the clustering effect is evaluated by the Davies-Bouldin index, and the formula is as follows:

[0037]

[0038] Among them, k is the number of clusters in the clustering, S i represents the within-cluster scatter of the i-th cluster, and M ij represents the centroid distance between the i-th and j-th clusters, and they are respectively expressed as:

[0039]

[0040] M ij = ||c i -c j ||

[0041] where C i represents the sample set of the i-th cluster, and c i represents the centroid of the i-th cluster; the smaller the value of DBI, the better the clustering effect.

[0042] Furthermore, in step S5, during the process of matching the fingerprint feature of the point to be located with the fingerprint features of the clustered fingerprint data, the Mahalanobis distance is used to measure the distance between two fingerprint vectors, and the formula is as follows:

[0043]

[0044] where x represents the n-dimensional vector to be calculated, μ represents the mean vector of the distribution, and S represents the covariance matrix of the distribution;

[0045] Calculate the Mahalanobis distance between the fingerprint feature vector of the point to be located and the mean feature vectors of each clustering center in the fingerprint database, select the K with the smallest distance as the optimal match, and then take the average of the K positions and output the coordinates as the final positioning result.

[0046] The beneficial effects of the present invention are as follows:

[0047] The present invention fully exploits the high time-frequency resolution, multipath richness, and high sensitivity to spatial position exhibited by terahertz signals during indoor propagation. By extracting stable and discriminative channel feature parameters to construct a fingerprint database, the positioning accuracy and robustness are significantly improved. The high time-frequency resolution characteristics of the terahertz band greatly improve the accuracy of signal delay and frequency domain feature extraction. Combining the multipath angle distribution and small-scale parameters such as delay spread obtained by a directional scanning high-gain horn antenna, the spatial channel fingerprint can be accurately characterized. The large-scale parameters such as the maximum power and path loss extracted by physical modeling and statistical methods further enhance the spatial discrimination. The weighted K-nearest neighbor algorithm combined with Mahalanobis distance matching eliminates the scale differences between different parameter dimensions, reducing the positioning error to the centimeter level.

[0048] Compared with existing communication technologies such as Wi-Fi, Bluetooth, or UWB, the present invention has stronger anti-multipath interference ability and environmental adaptability in complex indoor environments, does not require additional modification of the existing scenario, is applicable to various types of intelligent indoor spaces, can provide effective support for high-precision and high-reliability positioning services, and has broad application prospects and promotion value.

[0049] Other advantages, objects, and features of the present invention will be set forth in part in the following description, and in part will be obvious to those skilled in the art upon examination of the following, or may be learned by practice of the present invention. The objects and other advantages of the present invention may be realized and obtained by the following description. Description of the Drawings

[0050] In order to make the objectives, technical solutions, and advantages of the present invention more clear, the present invention will be described in detail preferably with reference to the accompanying drawings, where:

[0051] Figure 1 It is the overall flowchart of the indoor fingerprint positioning method based on terahertz channel characteristics in the embodiment of the present invention;

[0052] Figure 2 It is the flowchart of the DBSCAN algorithm in the embodiment of the present invention;

[0053] Figure 3 It is the schematic diagram of the positioning result example in the embodiment of the present invention. Detailed Embodiments

[0054] The following uses specific specific examples to illustrate the implementation manners of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the diagrams provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner. Without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0055] Among them, the accompanying drawings are only for illustrative purposes, showing only schematic diagrams, not physical diagrams, and should not be construed as a limitation to the present invention; in order to better illustrate the embodiments of the present invention, some components in the accompanying drawings will be omitted, enlarged, or reduced, and do not represent the dimensions of actual products; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the accompanying drawings may be omitted.

[0056] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "front", "rear", etc. indicating the orientation or positional relationship, they are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, the terms describing the positional relationship in the accompanying drawings are only for illustrative purposes and cannot be construed as a limitation of the present invention. For those of ordinary skill in the art, the specific meanings of the above terms can be understood according to specific circumstances.

[0057] Please refer to Figures 1 to 3 , which is an indoor fingerprint positioning method based on terahertz channel characteristics.

[0058] Embodiment

[0059] This embodiment first provides a detailed implementation manner of an indoor fingerprint positioning method based on terahertz channel characteristics. As shown in the flowchart Figure 1 shown, the method includes the following steps:

[0060] Step 1: Divide the area to be used for positioning into grids, establish a coordinate system, and record the coordinate information of each grid point position;

[0061] Step 2: Use a professional channel measurement device to measure the channel data of each point and save the data;

[0062] Step 3: Data preprocessing: Process the data through physical and statistical methods to obtain channel characteristics such as large-scale parameters and small-scale parameters in this environment, and then combine the obtained channel characteristics with coordinate information to construct a multi-dimensional fingerprint database;

[0063] Step 4: Use the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) clustering algorithm to cluster the database to form a clustered fingerprint database, so as to reduce the workload and time complexity of the subsequent matching process;

[0064] Step 5: Obtain the fingerprint features of the point to be located, use the Weighted K-Nearest Neighbors (WKNN) matching algorithm to input the fingerprint features of the point to be located into the clustered fingerprint database for matching, and output the two-dimensional coordinate values;

[0065] In step 1 of this embodiment, the specific process of establishing the coordinate system is as follows: The coordinate origin is located at the corner position in the southeast direction of the room, and the interval between coordinates is 0.8 m.

[0066] In step 2 of this embodiment, it specifically includes the following steps:

[0067] Step 21: Use a professional channel measurement device to measure the channel data at each point and establish an environment-based channel data set; the terahertz channel measurement system adopts a frequency-domain measurement method based on a vector network analyzer (VNA). It consists of a transmitter (Tx), a receiver (Rx), two radio frequency front ends (RF), a signal generator, and a VNA. Measure the divided grid points and randomly measure the points to be located. Place the transmitter Tx at the coordinate origin and the receiver Rx at the grid points and the points to be located to obtain two data sets, one is the data of the grid points and the other is the data of the points to be located.

[0068] In step 22: The VNA used in channel measurement supports a frequency range of 10 MHz - 50 GHz. Both the Tx and Rx antennas are linearly polarized horn antennas with a frequency range of 170 - 260 GHz and are both equipped with standard waveguides; the frequency range of the two-way power divider is 0.5 - 26.5 GHz. At the transmitting end, the local oscillation (LO) signal of 13.104 GHz generated by the signal generator is multiplied by 16 through a frequency multiplier to 210 GHz, and then mixed with the intermediate frequency (IF) signal of 5 - 15 GHz generated by the VNA to the 215 - 225 GHz frequency band; then the terahertz signal is transmitted and received by the horn antenna.

[0069] At the receiving end, generate the same LO signal (13.104 GHz), the terahertz signal is down-converted to an IF signal of 5 - 15 GHz and sent back to port 2 of the VNA to obtain the S21 parameter; the S21 parameter refers to the complex amplitude ratio (including amplitude and phase) when the signal enters from port 1 and exits from port 2, and is used to describe the forward transmission characteristics of the device under test. Among them, the output power of the signal generator is 0 dBm, and the output power of the VNA is less than or equal to -20 dBm. The signal measurement bandwidth B w is 10 GHz, so the time delay resolution Δτ = 1 / B w is 0.1 ns, and the path length resolution ΔL = c·Δτ is 3 cm, that is, multipath components with a path length difference greater than 3 cm can be separated. In addition, set the sweep frequency interval Δf to 5 MHz, then the number of sweep frequency points N = B W / Δf is 2001, and the maximum remaining time delay (the maximum delay of the MPC) τ m = 1 / Δf is 200 ns, so the maximum path length Lm = c·τ m is 60 m.

[0070] Step 23: The directional horn antenna adopted at the Tx end has a gain of 10 dBi, the half-power beamwidth (HPBW) in the E-plane of the antenna is 49°, and the HPBW in the H-plane is 46°; the directional horn antenna adopted at the Rx end has a gain of 25 dBi, the HPBW in the E-plane of the antenna is 8°, and the HPBW in the H-plane is 9°. Both the transmitting and receiving modules are installed on a customized multi-dimensional turntable, which can perform three-dimensional spherical scanning on the measurement scenario. The H-plane covers 360°, the E-plane covers 180°, and the minimum resolution is 1°. The Tx antenna is 1.6 m above the ground, and the Rx antenna is 1.51 m above the ground. During measurement, the Tx remains stationary, while the Rx rotates within 0° - 350° in the H-plane and -20° - 20° in the E-plane, receiving multipath components at a step size of 10°, which corresponds to the HPBW of the Rx antenna, to achieve omnidirectional coverage. The measurement parameters of the system are set as shown in Table 1.

[0071] Table 1

[0072]

[0073] Step 24: Since the S21 parameter recorded in the VNA for the measurement results contains the influences caused by the RF front-end, measurement system, etc., it is necessary to calibrate the vector network analyzer and the RF front-end in sequence before the measurement. The calibration of the vector network analyzer adopts the two-port short-circuit, open-circuit, load, and through (SOLT) method. The purpose of the RF front-end correction is to eliminate the influence of the transmitter and receiver RF front-ends on the measurement results, so as to obtain accurate and pure channel measurement data.

[0074] In step 3 of this embodiment, the collected data is processed through physical and statistical methods to obtain channel characteristics such as large-scale parameters and small-scale parameters in this environment. The extracted large-scale parameters include the maximum power P at each grid point m and the path loss PL, and the calculation methods are as follows:

[0075] PDAP(τ,θ RX ,φ RX ) = |h chann (τ,θ RX ,φ RX )| 2 (1)

[0076]

[0077] P m = 10lg(max(PDP(τ))) (3)

[0078] where τ represents the time delay, and h channel is the impulse response of the channel. The PDP is a basic channel parameter derived from the channel impulse response, aiming to characterize the power distribution characteristics of the signal in the time domain. In actual measurement, the directional scanning of a high-gain horn antenna is used to identify the multipath components in space, and the directional PDP is obtained, that is, the PDAP with the elevation angle of θ RX and the azimuth angle of φ RX .

[0079]

[0080] PL = G TX + g RX + L c + PL FS (5)

[0081] where PL FS represents the path loss in free space, d is the distance between the transceiver, f is the frequency, c is the speed of light, G TX and G RX are the transmitter gain and the receiver gain respectively, and L c is the loss of the cable.

[0082] The extracted small-scale parameters include the root mean square delay spread (RMSDS) and the root mean square angular spread (RMSAS) of each grid point. The RMSDS is represented by the second-order central moment of the PDP, and the calculation method is as follows:

[0083]

[0084] where τ l represents the time delay of the l-th multipath, and ψ l represents the direction of arrival (DOA) of the l-th multipath, which includes the azimuth angle of arrival (AOA) and the elevation angle of arrival (EOA).

[0085] The process of combining the extracted large-scale parameters and small-scale parameters with the coordinate information to construct a multi-dimensional fingerprint database is as follows: Each row in the database represents the channel characteristics of a measurement point. The first four columns are feature vectors: the first column is the maximum power, the second column is the path loss, the third column is the delay spread, and the fourth column is the angular spread; the last two columns are the label values, which are the X coordinate and the Y coordinate respectively, jointly constructing the multi-dimensional fingerprint database.

[0086] In step 4 of this embodiment, as Figure 2 shown, it is the flowchart of the DBSCAN algorithm. The DBSCAN clustering algorithm is used to cluster the database to form a clustered fingerprint database, dividing it into multiple small regions to reduce the workload and time complexity of the subsequent matching process. The steps are as follows:

[0087] First, initialize by setting two parameters required for clustering, the neighborhood radius ε and the number of points Minpts included in a cluster, and input the database D to be clustered; then create initial variables: the set of core objects The number of clustering clusters M = 0, the unvisited set Γ = D, and the clustering result Next, calculate the neighborhood subsample sets of all samples to obtain the set of core objects. If Γ, then perform cyclic clustering. Randomly select a core object o from the unvisited sample set and remove it. Initialize the queue Q = <o>; Further expand the current cluster and then generate new clustering clusters; finally, determine whether the core objects are exhausted. If so, output the final clustering result; otherwise, return to the loop for clustering.

[0088] By adjusting the neighborhood radius ε and the number of points Minpts included in a cluster, the optimal parameters are sought, and the clustering effect is evaluated by the Davies-Bouldin index (DBI). The formula is as follows:

[0089]

[0090] where k is the number of clusters in the clustering, S i represents the within-class scatter of the i-th cluster, and M ij represents the centroid distance between the i-th and j-th clusters.

[0091]

[0092] M ij = ||c i - c j || (10) where C i represents the sample set of the i-th cluster, and c i represents the centroid of the i-th cluster. The smaller the value of DBI, the better the clustering effect.

[0093] In step 5 of this embodiment, the fingerprint features of the point to be located are obtained, and the fingerprint features of the point to be located are input into the clustering fingerprint database for matching using the WKNN matching algorithm, and the two-dimensional coordinate values are output; during the matching process, the Mahalanobis distance is used to measure the distance between two fingerprint vectors to eliminate the scale difference between different dimensional parameters. The formula is as follows:

[0094]

[0095] where x represents the n-dimensional vector to be calculated, μ represents the mean vector of the distribution, and S represents the covariance matrix of the distribution. Calculate the Mahalanobis distance between the fingerprint feature vector of the point to be located and the feature mean vector of each clustering center in the fingerprint database, select the K with the smallest distance as the optimal match, and then take the average of the K positions and output the coordinates as the final positioning result.

[0096] As Figure 3 shown, it is the real position and estimated position of the indoor positioning object in this embodiment. It can be seen that the positioning error is only 0.11 m, and high-precision positioning can be achieved.

[0097] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the present technical solution, and they should all be covered within the scope of the claims of the present invention.< / o>

Claims

1. An indoor fingerprint positioning method based on terahertz channel characteristics, characterized in that: The method includes the following steps: S1. Divide the area to be located into grids, establish a coordinate system, and record the coordinate information of each grid point; S2. Obtain the terahertz channel data of each point through a terahertz channel measurement device and save the data; S3. Use physical and statistical methods to process the terahertz channel data set to obtain terahertz channel characteristics; combine the terahertz channel characteristics with the coordinate information to construct a multi-dimensional fingerprint database; S4. Use the density-based spatial clustering of applications with noise (DBSCAN) algorithm to cluster the multi-dimensional fingerprint database to obtain a clustered fingerprint database; S5. Obtain the fingerprint features of the point to be located, use the weighted K-nearest neighbor matching algorithm to match the fingerprint features of the point to be located with the fingerprint features in the clustered fingerprint database, and output the two-dimensional coordinate values of the matching result to complete the positioning.

2. The indoor fingerprint positioning method based on terahertz channel characteristics according to claim 1, characterized in that: In step S1, the process of establishing the coordinate system is as follows: Select an indoor location as the coordinate origin, set the axis directions and the distance between each coordinate point.

3. The indoor fingerprint positioning method based on terahertz channel characteristics according to claim 1, characterized in that: In step S2, the terahertz channel measurement device measures the grid points divided by the frequency domain measurement method based on a vector network analyzer; among them, the terahertz channel measurement device includes a transmitter Tx, a receiver Rx, two radio frequency front ends RF, a signal generator, a two-way power divider, and a vector network analyzer VNA.

4. The indoor fingerprint positioning method based on terahertz channel characteristics according to claim 3, characterized in that: The process by which the terahertz channel measurement device obtains and saves the terahertz channel data includes: S21. Place the transmitter Tx at the coordinate origin of the area to be located, and randomly place the two receivers Rx at the divided grid points and the point to be located to obtain the terahertz channel data set of the grid points and the terahertz channel data set of the point to be located; S22. Amplify the local oscillator signal generated by the signal generator of the transmitter Tx to a preset frequency through a frequency multiplier, mix it with the intermediate frequency signal generated by the vector network analyzer VNA to a preset frequency band, and then the terahertz signal is transmitted by the horn antenna of the transmitter and received by the horn antenna of the receiver; at the same time, the signal generator of the receiver Rx generates a local oscillator signal, and down-converts the terahertz signal to an intermediate frequency signal of a preset frequency; send the intermediate frequency signal back to the vector network analyzer VNA to obtain the complex amplitude ratio parameter, which refers to the complex amplitude ratio when the signal is incident from the port and when it is output from the port, and is used to describe the forward transmission characteristics of the device under test; S23. Install the transmitter Tx and the receiver Rx on a multi-dimensional turntable in a specific manner, where the installation is completed with the half-power beam widths of the directional horn antennas of the transmitter Tx and the receiver Rx as the installation adjustment parameters; at the start of the measurement, the transmitter Tx remains stationary, and the receiver Rx rotates within a preset range in the H plane and the E plane and receives multipath components with a preset step size to complete the measurement.

5. The indoor fingerprint positioning method based on terahertz channel characteristics according to claim 4, wherein: During the measurement process by the terahertz channel measurement device, its parameter configuration includes: the center frequency of the terahertz measurement system, the local oscillator signal frequency of the signal generator, the frequency multiplication factor of the frequency multiplier, the intermediate frequency signal frequency generated by the VNA, the start frequency, the stop frequency and the bandwidth range, the intermediate frequency bandwidth, the number of sweep points, the maximum power of IF IN, the transmitter antenna gain, the receiver antenna gain, the transmitter half-power beam width, the receiver half-power beam width, the delay resolution, the maximum remaining delay of the multipath component, the receiver azimuth range and elevation range, the sweep interval, and the maximum power of LO IN.

6. The indoor fingerprint positioning method based on terahertz channel characteristics according to claim 1, wherein: In step S3, the obtained terahertz channel characteristic parameters include large-scale parameters and small-scale parameters based on the environment, where the large-scale parameters include the maximum power P of each grid point m and path loss PL; the small-scale parameters include the root mean square delay spread RMSDS and the root mean square angular spread RMSAS of each grid point; The process of combining the extracted large-scale parameters and small-scale parameters with coordinate information to construct a multi-dimensional fingerprint database is as follows: Each row in the database represents the channel characteristics of a measurement point, where each column represents a feature vector or a label value. Among them, the first column is the maximum power, the second column is the path loss, the third column is the delay spread, and the fourth column is the angular spread; the last two columns are label values, which are the X coordinate and the Y coordinate respectively, jointly constructing the multi-dimensional fingerprint database.

7. The indoor fingerprint positioning method based on terahertz channel characteristics according to claim 6, wherein: The maximum power P at each grid point of the large-scale parameter m is calculated as follows: PDAP(τ,θ RX ,φ RX )=|h chann (τ,θ RX ,φ RX )| 2 P m = 10 lg(max(PDP(τ))) Among them, τ represents the time delay, and PDP(·) represents the basic channel parameter derived from the channel impulse response for characterizing the power distribution characteristics of the signal in the time domain; θ RX is the elevation angle of the horn antenna, and φ RX is the azimuth angle of the horn antenna; h cha (·) represents the impulse response of the channel; The calculation method of the path loss PL at each grid point of the large-scale parameters is as follows: PL = G TX +G RX +L c +PL FS Among them, PL FS represents the path loss in free space, d is the distance between the transceiver, f is the frequency, c is the speed of light, G TX and G RX are the gains of the transmitter and the receiver respectively, and L c is the loss of the cable; The root mean square delay spread RMSDS at each grid point of the small-scale parameters is represented by the second central moment of PDP(·), and the calculation method is as follows: The calculation method of the root mean square angular spread RMSAS at each grid point of the small-scale parameters is as follows: where τ l represents the time delay of the l-th multipath, and ψ l represents the direction of arrival of the l-th multipath, which includes the azimuth angle of arrival and the elevation angle of arrival.

8. A method for indoor fingerprint positioning based on terahertz channel characteristics according to claim 1, characterized in that: In step S4, after clustering the database into a clustering fingerprint database by the DBSCAN clustering algorithm, the database is divided into multiple small regions, and the process is as follows: By adjusting the neighborhood radius ε and the number of points Minpts included in a cluster to find the optimal parameters, the clustering effect is evaluated by the Davies-Bouldin index, and the formula is as follows: where k is the number of clusters, and S i represents the within-class scatter of the i-th cluster, and M ij represents the centroid distance between the i-th and j-th clusters, which are respectively expressed as: M ij = ||c i -c j || Among them, C i represents the sample set of the i-th cluster, and c i represents the centroid of the u-th cluster; the smaller the value of DBI, the better the clustering effect.

9. The indoor fingerprint positioning method based on terahertz channel characteristics according to claim 1, wherein: In step S5, during the process of matching the fingerprint features of the point to be located with the fingerprint features of the clustering fingerprint data, the Mahalanobis distance is used to measure the distance between two fingerprint vectors, and the formula is as follows: Among them, x represents the n-dimensional vector to be calculated, μ represents the mean vector of the distribution, and S represents the covariance matrix of the distribution; Calculate the Mahalanobis distance between the fingerprint feature vector of the point to be located and the feature mean vector of each clustering center in the fingerprint database, select the K with the smallest distance as the optimal match, and then take the average of the K positions and output the coordinates as the final positioning result.

Images