A carrier phase localization method based on node selection in a dynamic topology environment

Abstract

This invention proposes a carrier phase positioning method based on node selection in a dynamic topology environment, applicable to fields such as communication and positioning. This method eliminates clock bias through differential calculations in the terminal domain and access point domain, and improves positioning accuracy using GDOP as a node selection criterion. A time variable from the dynamic topology scenario is introduced into the carrier phase measurement equation; and the high-precision position change estimate calculated from the carrier phase equation is fused with the unbiased TDoA positioning result to obtain a relatively high-precision coarse position estimate. These coarse estimates are then used to linearize and expand the phase measurement equation, achieving high-precision integer ambiguity resolution and high-precision positioning.

Description

Technical Field

[0001] This invention relates to the fields of communication and positioning, and in particular to the positioning problem based on carrier phase in 5G NR. Background Technology

[0002] Location-based services originated from the Global Positioning System (GPS). The navigation and positioning functions provided by the Global Navigation Satellite System (GNSS) are of paramount importance for commercial, public safety, and military applications. In open outdoor environments, GNSS systems can provide stable and reliable high-precision positioning services. However, with the development of mobile communication technology, in complex environments, GNSS signals are severely interfered with by buildings, walls, and other objects, leading to decreased positioning accuracy or even failure. The demand for location-based services can no longer be met by GNSS alone. The need for high-precision positioning is constantly emerging, thus requiring new solutions and technologies to address the problem of high-precision positioning in obstructed environments. In 5G communication systems, positioning performance has been comprehensively improved, providing an effective solution to the positioning problem.

[0003] There have been many research achievements in wireless positioning technology, which can be mainly divided into range-based positioning technology, angle-based positioning technology, and signal strength-based positioning technology.

[0004] The most classic algorithm for range-based positioning is ToA (Time of Arrival), which transforms ToA measurements into distance estimations and then solves them using nonlinear geometric equations. Another technique is TDoA (Time-of-Arrival), which establishes a hyperbolic equation based on distance difference measurements, reducing the impact of clock asynchrony between the terminal and nodes. Angle-based positioning requires measuring the angle of arrival (OA) between the terminal and the access points, determining the target location based on the intersection of the incident angles from each access point. Signal strength-based positioning utilizes channel fading models, estimating the location of the terminal through signal attenuation; fingerprint positioning is a common example.

[0005] Among range-based positioning technologies, ToA and TDoA positioning are the most widely used. However, limitations in equipment performance and time delay measurement errors caused by the wireless propagation environment prevent the acquisition of high-precision distance measurements, thus affecting positioning accuracy. The existence of clock skew between systems further reduces positioning accuracy. Angle-based positioning technologies require high-precision angle estimation, necessitating large-scale antenna arrays. Therefore, while this method has lower requirements for time synchronization within the system, it places high demands on the equipment. Furthermore, when the target terminal is far from the base station, even small measurement errors can lead to significant positioning errors. Signal strength-based positioning technologies, such as fingerprint positioning, require prior measurement of the scene to establish a database and rely on the accuracy of the channel attenuation model. When the wireless environment changes, the database needs to be rebuilt, resulting in high deployment and time costs, and difficulty in guaranteeing high positioning accuracy.

[0006] Traditional distance measurement methods often rely on signal propagation delay or signal attenuation, resulting in significant ranging errors. However, by utilizing the high-carrier-frequency phase measurement signals in 5G NR, centimeter-level ranging accuracy can be achieved. Therefore, carrier phase positioning technology based on 5G NR can achieve considerably high positioning accuracy.

[0007] The method of this invention is mainly aimed at the positioning process in dynamic topology environment. Based on node optimization and carrier phase and time delay measurement equations, it dynamically selects access points to ensure optimal geometric distribution, obtains high-precision position estimation through fusion algorithm, and linearizes the phase measurement equations at multiple time points to solve for high-precision integer ambiguity and position. Summary of the Invention

[0008] The technical problem to be solved by this invention is: for the dynamic topology environment in 5G NR, a carrier phase positioning method based on node selection in the dynamic topology environment is proposed. This method selects nodes based on the geometric dilution of precision (GDOP), and on this basis, the time variable in the dynamic topology is introduced into the carrier phase measurement equation, and linearized to calculate the integer ambiguity and high-precision estimation of the position, which can achieve centimeter-level high-precision positioning in the dynamic topology environment.

[0009] The technical solution adopted in this invention is as follows:

[0010] A carrier phase localization method based on node selection in a dynamic topology environment includes the following steps:

[0011] (1) The terminal to be located roughly calculates the current position estimate from the TDoA measurement value obtained at the current time, and calculates the GDOP value of all access point combinations. The access point combination with the smallest GDOP value is selected as the access point to participate in the positioning algorithm, and the TDoA positioning result is calculated.

[0012] (2) Introduce the time variables in the dynamic topology environment into the carrier phase measurement equation and perform differential processing in the time domain to obtain the high-precision relative position estimate change;

[0013] (3) The obtained high-precision relative position estimate change is fused with the TDoA positioning result to obtain a high-precision coarse position estimate;

[0014] (4) Using the high-precision coarse position estimate as the Taylor expansion point, the carrier phase measurement equations at multiple times are linearized and expanded. The equations at multiple times are solved simultaneously to obtain the double-difference integer ambiguity, which is then converted into the single-difference integer ambiguity. This yields the high-precision distance difference estimation result, and finally, the high-precision position estimate is obtained by the TDoA algorithm.

[0015] Furthermore, in step (1), the approximate location of the terminal to be located is first obtained from the TDoA measurement value obtained at the current time. Based on the approximate location, the GDOP value of each access point combination is calculated, and the access point combination with the smallest GDOP value is obtained by the following method:

[0016] When the terminal is at one of M access points m1,...,m M When within the communication range, the distance difference measurement value is used.

[0017] r (i,1) =r (i) -r (1)

[0018] Where, r (i) Let r be the physical distance between the terminal and access point i, where i = 2, ..., M. (1) r is the physical distance between the terminal and the reference access point l. (i,1) This is the difference between the physical distance between the terminal and access point i and the physical distance between the terminal and access point 1;

[0019] Rearranging the total differential with respect to all access points into matrix form, we have:

[0020]

[0021] Matrix C represents the geometric distribution relationship between the terminal and the access point;

[0022] The positioning error [dx dy dz] is obtained using the weighted least squares method. T covariance matrix

[0023] P = [C T W -1 C] -1

[0024] Therefore, the final expression for GDOP is:

[0025]

[0026] Where W is the covariance matrix of the distance difference measurements;

[0027] Select N access points n1,...,n N For location services, N ≤ M; the access point closest to the terminal is taken as the reference access point n1, and other access points are selected based on the following criteria.

[0028]

[0029] Furthermore, the specific process of step (3) is as follows:

[0030] Step 3-1: At the initial time t=0, use the TDoA positioning result ξ0 as the initial value of the fused positioning result.

[0031] Step 3-2: At time t≥1, estimate the change Δv based on the high-precision relative position. t,t-1 Fusion localization results at time t-1 The carrier phase positioning result at time t Updated to

[0032] Step 3-3: Obtain the TDoA location result ξ at the current moment. t Carrier phase positioning results The final fusion positioning result is obtained through fusion.

[0033]

[0034] Among them, w TDoA and w CARR These are the weights of the TDoA positioning result and the carrier phase positioning result, respectively, and their sum is 1;

[0035] Step 3-4: Complete the calculation for the current moment, and return to step 3-2 for the next moment.

[0036] Compared with the prior art, the present invention has the following advantages:

[0037] The node optimization scheme based on GDOP proposed in this invention can ensure that the geometric layout is always reasonable in a dynamic topology environment and that the positioning performance is not limited by the node layout. The high-precision positioning method based on carrier phase introduces the time variable in the dynamic topology, which can overcome the influence of the dynamic movement of nodes and terminals and achieve high-precision positioning. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of a carrier phase positioning method based on node selection in a dynamic topology environment according to the present invention.

[0039] Figure 2 This is a schematic diagram illustrating the impact of the node layout on positioning performance in this invention.

[0040] Figure 3 This is a schematic diagram illustrating the accuracy of relative position estimation and absolute position estimation in this invention. Detailed Implementation

[0041] The following is in conjunction with the appendix Figures 1-3 The present invention will now be described in further detail.

[0042] like Figure 1 As shown, a carrier phase positioning method based on node selection in a dynamic topology environment has the following specific steps:

[0043] Step (1) considers a positioning scenario with multiple wireless access points, where both the terminal and the access points are in motion. Since the layout of nodes has a significant impact on positioning accuracy, it is necessary to dynamically select the optimal node combination during the positioning process to achieve the best positioning performance. Figure 2 The impact of node layout on positioning accuracy is shown; the smaller the GDOP value, the higher the positioning accuracy.

[0044] In the field of positioning, the Geometrical Dilution of Precision (GDOP) is commonly used to describe the impact of node geometry on positioning performance. It is defined as follows:

[0045]

[0046] Where P is the covariance matrix of the positioning error, σ x σ y σ z These represent the positioning errors in the three directions. A smaller GDOP indicates that the corresponding layout has higher positioning accuracy under this measurement error.

[0047] When the positioning model uses distance difference measurements, there is

[0048] r (i,1) =r (i) -r(1)

[0049] Where, r (i) Let r be the physical distance between the terminal and access point i, where i = 2, ..., M. (1) r is the physical distance between the terminal and the reference access point l. (i,1) This is the difference between the physical distance between the terminal and access point i and the physical distance between the terminal and access point 1;

[0050] Finding the total differential of the above equation, we can obtain

[0051]

[0052] Among them, (a (i) ,b (i) ,c (i) (x, y, z) represents the location coordinates of the i-th access point, and (x, y, z) represents the location of the terminal.

[0053] Rearranging the total differential with respect to all access points into matrix form, we have:

[0054]

[0055] Matrix C represents the geometric distribution relationship between the terminal and the access point, and its specific expression is as follows:

[0056]

[0057] Among them, R i This represents the distance between the i-th access point and the terminal, where... This represents the position coordinates of the i-th wireless access point at time t.

[0058] Using the weighted least squares method, the positioning error [dx dy dz] can be obtained. T covariance matrix

[0059] P = [C T W -1 C] -1

[0060] Therefore, the final expression for GDOP is:

[0061]

[0062] Where W is the covariance matrix of the distance difference measurements, and tr() is the trace operation of the matrix.

[0063] When the terminal is at one of M access points m1,...,m M When the communication range is within the specified range, N (N < M) access points n1,...,n need to be selected. NParticipate in positioning. The access point closest to the terminal is selected as the reference access point n1. Other access points are selected based on the following indicators.

[0064]

[0065] Before the algorithm starts, the coarse location estimate of the terminal (such as the TDoA positioning result) is substituted into the above formula to select the node combination corresponding to the minimum GDOP value to participate in the algorithm calculation and obtain the TDoA positioning result.

[0066] Step (2): Assume there are M wireless access points and one reference terminal participating in the positioning process, and during the positioning process, the trajectory of the moving wireless access points... The locations of the reference terminal r are known, but there is an unknown clock difference between the access point and the terminal s to be located. By taking access point 1 as the reference access point, the TDoA measurement value can be obtained.

[0067]

[0068] Where Δ() (i,1) This represents a difference operation, where c represents the speed of electromagnetic wave propagation in space. This represents the difference in distance between the terminal s to be located and the i-th access point and the first access point at time t. This represents the clock offset between the terminal s to be located and the i-th access point and the first access point at time t. It represents the difference in measurement noise between the terminal to be located and the i-th access point and the 1-th access point at time t.

[0069] To eliminate the unknown clock bias term, the TDoA measurement at the reference terminal is introduced.

[0070]

[0071] in, This represents the difference in distance between the terminal r to be located and the i-th access point and the first access point at time t. This represents the clock offset between the terminal r to be located and the i-th access point and the first access point at time t. Let represent the difference in measurement noise between the terminal r to be located and the i-th access point and the first access point at time t. Dividing the TDoA measurement differences between the two terminals, we have...

[0072]

[0073] Due to clock skew Both represent the clock difference between access point i and the reference access point, thus eliminating the time-varying clock difference term after terminal differential. Moving the known terms in the above equation to the left side of the equal sign yields the TDoA measurement value after clock skew elimination.

[0074]

[0075] in, express and The difference results between them

[0076] Similar to the processing of TDoA measurements, after differentiating the phase measurements in the terminal domain and access point domain, we can obtain...

[0077]

[0078]

[0079]

[0080]

[0081]

[0082] Where λ is the wavelength of the signal. Let represent the phase values ​​from the i-th wireless access point measured by terminals s and r at time t, respectively. This represents the difference between the phase values ​​measured by the terminal s to be located from the i-th access point and the 1-th access point at time t. This represents the difference between the phase values ​​measured by the terminal s to be located at time t, which are from the i-th access point and the 1-th access point. Let S and R represent the integer ambiguity of the corresponding access point i when terminals s and r are initially locked, respectively. This represents the difference in integer ambiguity between the i-th and 1st access points for the terminal. Let represent the phase measurement errors between terminals s and r and access point i, respectively, and assume that their variances are . Gaussian distribution, These represent the phase measurement noise differences between terminals s and r and access points i and 1, respectively.

[0083] Through the above process, the TDoA measurement value and carrier phase measurement value with constant deviation elimination are finally obtained, which helps to improve the positioning accuracy of the algorithm. Figure 3 This demonstrates that the position change estimate obtained by this algorithm has high accuracy.

[0084] At the initial moment, the initial position of the mobile terminal is obtained first from the TDoA measurement value. And find the geometrically optimal node combination. Perform time-difference on the carrier phase measurement equation of mobile terminal s to obtain...

[0085]

[0086] in This represents the phase measurement value between terminal s and access point i at time t; This represents the distance between terminal s and access point i at time t; f represents the phase measurement noise between terminal s and access point i at time t; c Indicates the carrier frequency of the phase measurement signal; This represents the clock offset between terminal s and access point i at time t.

[0087] Since integer ambiguity remains unchanged after the phase-locked loop is locked, it is eliminated.

[0088] For the reference terminal, there are

[0089]

[0090] in This represents the phase measurement value between terminal r and access point i at time t; This represents the distance between terminal r and access point i at time t; This represents the phase measurement noise between terminal r and access point i at time t; This represents the clock offset between terminal r and access point i at time t.

[0091] remember Let be the change in distance between the fixed terminal and access point i between time t-1 and t. This quantity is known when the base station trajectory is known.

[0092] To eliminate the clock bias term, equations (1) and (2) are differentiated in the terminal domain and access point domain, resulting in...

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099] in, The difference between access point i and the reference access point in terms of distance to the reference terminal varies with the movement trajectory of the access point.

[0100] Move all known and observed quantities to the left side of the equation, and... Substituting, we can get

[0101]

[0102]

[0103]

[0104]

[0105]

[0106]

[0107]

[0108]

[0109]

[0110] Among them, (x t ,y t ,z t ) represents the position v to be estimated. t , λ represents the position of the i-th access point at time t, and λ represents the wavelength of the phase measurement signal.

[0111] The above formula includes two positional quantities v t and v t-1 Therefore, given the position v of the previous time step t-1 Then the positions at subsequent time steps can be iteratively calculated. For all access points i = 2, 3, ..., M, writing the above equation in matrix form, we have...

[0112] C t z t =q t

[0113] in,

[0114]

[0115] By using the least squares method, we can obtain v t The estimated value

[0116]

[0117] in[] (1:3) This represents a subvector consisting of the first to third elements of the vector []. This estimate is biased by the initial position, but the time-domain differencing eliminates the influence of integer ambiguity on the solution result; therefore, this algorithm has high estimation accuracy for position changes. The estimated position change is...

[0118]

[0119] Step (3), in order to eliminate the initial position To investigate the impact of bias on the iteration results, the unbiasedness of TDoA measurements is utilized to fuse them with high-precision positional changes, resulting in a more accurate fused positioning result. This can reduce the error in subsequent Taylor expansion. The specific steps are as follows:

[0120] Initialization: At the initial time t=0, the TDoA positioning result ξ0 is obtained from the TDoA measurement value with time-varying clock error elimination, and ξ0 is taken as the initial value of the fused positioning result.

[0121] Step 3-1: At time t≥1, obtain the carrier phase positioning result from the carrier phase position change estimation equation. Then calculate the change in position Δv t,t-1 ;

[0122] Step 3-2: Estimate the change Δv based on the high-precision position. t,t-1 Fusion localization results at time t-1 Carrier phase positioning results Updated to

[0123] Step 3-3: Calculate the TDoA positioning result ξ at the current moment from the TDoA measurement value after time-varying clock error elimination. t and fusion ξ t and The final fusion localization result is obtained.

[0124]

[0125] Among them, w TDoA and w CARR These are the weights of the TDoA positioning result and the carrier phase positioning result, respectively, and their sum is 1.

[0126] Steps 3-4: Complete the calculation for the current moment. Execute step 1 at the next moment.

[0127] Step (4) uses the relatively accurate coarse position estimate as the Taylor expansion point to linearize the carrier phase measurement equation, which reduces nonlinear errors and facilitates accurate integer ambiguity resolution. From the terminal domain differential phase measurement equation, Taylor expansion is performed while retaining the first-order terms, resulting in...

[0128]

[0129] in,

[0130]

[0131]

[0132]

[0133] remember To eliminate the clock skew term, subtracting the above equation with respect to the access point yields:

[0134]

[0135]

[0136]

[0137]

[0138]

[0139]

[0140] Establish a system of equations for all access points.

[0141] ψ t =H t x t -2πN+γ t

[0142] in,

[0143]

[0144] x t =(x t ,y t ,z t ) T ,

[0145]

[0146] N=[ΔN (2,1) ΔN (3,1) ... ΔN (M,1) ] T,

[0147]

[0148] Since this system of equations is an indeterminate system, and considering that the integer ambiguity does not change with time, a unique solution can be obtained by simultaneously solving the system of equations at multiple time points. Simultaneously solving multiple systems of equations at K time points yields...

[0149] ψ=Hζ+γ

[0150] in,

[0151] ψ = [ψ1 ψ2 … ψ] K ] T , ζ=[x1 … x K N] T , γ=[γ1 γ2 … γ K ] T

[0152]

[0153] Matrix I is an M-1 order identity matrix.

[0154] The approximate solution to the system of equations is obtained by using the weighted least squares method. for

[0155]

[0156] Among them, Q ψ For measured values The covariance matrix.

[0157] Then, by using the approximate solution... The covariance matrix Q ζ The floating-point solution N can be obtained. float The covariance matrix Q NN

[0158]

[0159] Q xx Representing an approximate solution The covariance matrix at the midpoint, Q xN Q Nx Q represents the covariance matrix of position and integer ambiguity. NN N represents the integer ambiguity floating-point solution. float The covariance matrix.

[0160] Determine the floating-point N float and Q NN The fixed solution N for integer ambiguity N can be obtained by using it as input to the LAMBDA algorithm. fix .

[0161] According to the solution N fix The phase measurements of the terminal domain and access point domain differential are corrected to the phase measurements of the access point domain differential with clock skew eliminated.

[0162]

[0163] Measured values Clock skew has been eliminated, and accurate integer ambiguity estimates have been included, thus the distance difference derived from this phase measurement is accurate. It has high ranging accuracy, and finally, a high-precision position estimation can be achieved through a distance difference-based positioning algorithm.

Claims

1. A carrier phase positioning method based on node selection in a dynamic topology environment, characterized in that, Includes the following steps: (1) The terminal to be located roughly calculates the current position estimate from the TDoA measurement value obtained at the current time, and calculates the GDOP value of all access point combinations. The access point combination with the smallest GDOP value is selected as the access point to participate in the positioning algorithm, and the TDoA positioning result is calculated. (2) Introduce the time variables in the dynamic topology environment into the carrier phase measurement equation and perform differential processing in the time domain to obtain the high-precision relative position estimate change; (3) The obtained high-precision relative position estimate change is fused with the TDoA positioning result to obtain a high-precision coarse position estimate; (4) Using the high-precision coarse position estimate as the Taylor expansion point, the carrier phase measurement equations at multiple times are linearized and expanded. The equations at multiple times are solved simultaneously to obtain the double-difference integer ambiguity, which is then converted into the single-difference integer ambiguity. This yields the high-precision distance difference estimation result, and finally, the high-precision position estimate is obtained by the TDoA algorithm.

2. The carrier phase positioning method based on node selection in a dynamic topology environment according to claim 1, characterized in that, In step (1), the approximate location of the terminal to be located is first obtained from the TDoA measurement value obtained at the current time. Based on the approximate location, the GDOP value of each access point combination is calculated. The access point combination with the smallest GDOP value is obtained by the following method: When the terminal is at M access points m1,...,m M When within the communication range, the distance difference measurement value is used. r (i,1) ＝r (i) -r (1) Where, r (i) Let r be the physical distance between the terminal and access point i, where i = 2, ..., M. (1) r is the physical distance between the terminal and the reference access point l. (i,1) This is the difference between the physical distance between the terminal and access point i and the physical distance between the terminal and access point 1; Rearranging the total differential with respect to all access points into matrix form, we have: Matrix C represents the geometric distribution relationship between the terminal and the access point; The positioning error [dx dy dz] is obtained using the weighted least squares method. T covariance matrix P=[C T W -1 C] -1 Therefore, the final expression for GDOP is: Where W is the covariance matrix of the distance difference measurements; Select N access points n1,...,n N For location services, N ≤ M; the access point closest to the terminal is taken as the reference access point n1, and other access points are selected based on the following criteria.

3. The carrier phase positioning method based on node selection in a dynamic topology environment according to claim 1, characterized in that, The specific process of step (3) is as follows: Step 3-1: At the initial time t=0, use the TDoA positioning result ξ0 as the initial value of the fused positioning result. Step 3-2: At time t≥1, estimate the change Δv based on the high-precision relative position. t,t-1 Fusion localization results at time t-1 The carrier phase positioning result at time t Updated to Step 3-3: Obtain the TDoA location result ξ at the current moment. t Carrier phase positioning results The results are then merged to obtain the final fusion positioning result. Among them, w TDoA and w CARR These are the weights of the TDoA positioning result and the carrier phase positioning result, respectively, and their sum is 1; Step 3-4: Complete the calculation for the current moment, and return to step 3-2 for the next moment.

Images