Vehicle-mounted positioning system based on LTE signal

Abstract

The invention relates to the field of electronic communication and wireless positioning, in particular to a long-distance cellular positioning technology. The method aims at some special scenes such as tunnels, urban canyons and other complex terrains. In the above scene, the GNSS signal has the problems of positioning failure and poor positioning precision. Based on the mature construction of domestic commercial base stations at present and the realization of LTE signal coverage in most of areas, the method introduces cellular signals (LTE) as opportunity signals, completes channel estimation from the base stations to user equipment by using pilot signals in the cellular signals, extracts time delay parameters from an estimated channel matrix through an ESPRIT (Estimation Signal Random Interference Technology) algorithm, and finally achieves the coverage of LTE signals in the most areas. Therefore, TOA estimation is completed, and pseudo-range measurement is realized. After pseudo-range measurement is completed, the precision and robustness of the system are improved through extended Kalman filtering (EKF), and the stable and high-precision vehicle-mounted positioning system is successfully realized through simulation and actual measurement verification.

Description

Technical Field

[0001] The present invention designs a vehicle-mounted positioning system based on LTE signals in a complex GNSS environment and provides a hardware solution. Background Art

[0002] In recent years, the most widely used positioning system in the world is satellite positioning, such as the Global Positioning System (GPS) or the Beidou Global Positioning Navigation System. The GPS global positioning system uses satellites to navigate the ground. 98% of the area on the earth is covered by GPS. Today's GPS technology can provide high-precision positioning performance, speed measurement and timing information. The average positioning accuracy of GPS is generally around 10m.

[0003] However, GNSS navigation solutions will significantly degrade in the depths of urban canyons. Although the GNSS system has basically solved the outdoor positioning and navigation problems after more than half a century of development, the satellite signal transmission attenuates quickly in complex environments such as indoors and urban canyons, or is even completely blocked, making it difficult to provide high-performance, sustainable location services. Therefore, research on cellular data positioning (LTE / NR) to adapt to complex indoor and outdoor environments has become a new development direction for positioning and navigation.

[0004] When 2G / 3G networks became popular, cellular data could actually be used for positioning. In the early years, there were single-base station and multi-base station positioning methods. However, due to the narrow frequency band, low transmission rate, unstable network and high latency of 2G / 3G networks, the positioning based on 2G / 3G networks is often not accurate, usually more than 100 meters, and has little reference value. The arrival of 4G networks has greatly improved these shortcomings. The main advantages of LTE signals are as follows: the frequency bandwidth can reach 20MHz, the downlink data rate reaches 100Mbps, and the uplink data rate reaches 50Mbps. At the same time, the cell capacity is increased and the delay of the communication system is greatly reduced. These advantages allow LTE signals to have excellent performance in positioning. In addition, the latest 3GPP document also points out that the goal is to use cellular data to achieve a positioning accuracy of less than 3m indoors and less than 10m outdoors. There is no doubt that the prospect of using cellular data for high-precision positioning is very broad.

[0005] Combining multiple positioning technologies is of great significance in enhancing the robustness and environmental adaptability of the system and ensuring the high accuracy of navigation and positioning. The best estimation method used in precise positioning is Kalman Filtering (KF), and for systems with obvious nonlinearity such as navigation and positioning, the more widely used nonlinear estimation method is the Extended Kalman Filtering (EKF). The dynamic position estimation of fusion positioning can be achieved by using the extended Kalman filter, which can improve the positioning accuracy of the system and enhance the robustness and environmental adaptability of the system.

[0006] In the field of information theory and statistical inference, source estimation is an important task, especially when facing the number of unknown sources, the problem becomes more challenging. The MDL principle is a widely used theoretical framework that combines the ideas of Bayesian inference and entropy coding for data modeling and model selection. Its core idea is to find the model that can most concisely describe the data, that is, the balance between model complexity and the accuracy of fitting data, and find the minimum description length to predict the target variable and the transmission data model. In the problem studied in this paper, the MDL principle is used to estimate the number of multipath channels.

[0007] ESPRIT is a method for estimating signal parameters using rotation invariance techniques. There are two main types of ESPRIT algorithms: array-based and beam-based.

[0008] The prior art does not have a solution that combines the above methods for a vehicle positioning system based on LTE signals. Summary of the invention

[0009] The purpose of the present invention is to solve the problem of poor accuracy and low robustness of the vehicle-mounted positioning and navigation system in a complex GNSS environment, especially in a scenario where the GNSS signal is lost.

[0010] The technical solution is as follows:

[0011] A vehicle-mounted positioning system based on LTE signals, comprising the following modules: a cell search module, a frame synchronization module, a pseudo-range calculation module and an extended Kalman filter module;

[0012] Among them, the cell search module identifies base station signals near the area and distinguishes these base station signals;

[0013] The frame synchronization module uses PSS and SSS to complete the synchronization of LTE data frames and provides channel estimation to the pseudorange calculation module;

[0014] The pseudorange calculation module extracts pseudorange information from the LTE data frame and uses the MDL criterion and ESPRIT algorithm to extract the time delay from the base station to the receiver from the channel, thereby realizing pseudorange estimation;

[0015] The pseudorange estimation value obtained by the pseudorange calculation module is further processed by the extended Kalman filter module to reduce the error caused by the clock offset between the receiver and each base station, thereby outputting a smoother and more accurate position result.

[0016] In the present invention, we mainly estimate the pseudo-range from the base station to the receiver through LTE signals to achieve positioning, and after positioning, extend the Kalman filter to further improve the accuracy and robustness of the positioning system.

[0017] The entire vehicle positioning system mainly includes the following parts. First, the cell search module determines the cell ID of the base station corresponding to the received signal through the PSS (primary synchronization signal) and SSS (secondary synchronization signal) in LTE, and tracks the LTE data frame. Then, through the LTE signal demodulation module, the LTE data is demodulated and mapped to the corresponding RE (resource element), thereby extracting the required pilot symbol CRS (cell reference signal). Then, the extracted CRS signal is used through the pseudo-range calculation module, and the MDL criterion is used to determine the number of multipaths and the pseudo-range is determined through the ESPRIT algorithm, which effectively analyzes the multipath environment of the channel and improves the estimation accuracy of the LOS path delay. Then, the positioning is achieved through the TOA method, and finally the positioning accuracy is improved through the designed extended Kalman filter.

[0018] Beneficial effects of the present invention:

[0019] (1) Improve the robustness of the positioning system: In complex GNSS conditions, positioning can be achieved through cellular signals, namely LTE signals, which overcomes the problem of being unable to position in the absence of GNSS signals and provides a reliable solution for achieving positioning based on cellular signals.

[0020] (2) The accuracy of positioning through cellular signals is increased. First, the number of multipaths of the received signal is determined by the MDL criterion, and then the super-resolution algorithm ESPRIT is used to estimate the time delay of each path, so as to accurately obtain the pseudo-range from the receiver to the base station. The TOA method is then used to achieve positioning, and finally the designed extended Kalman filter is introduced to further improve the stability and accuracy of the entire positioning system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a structural diagram of the vehicle-mounted LTE positioning system of the present invention;

[0022] Figure 2 is a structural diagram of a cell search module of the present invention;

[0023] Figure 3 It is a structural diagram of the key positioning algorithm of the present invention;

[0024] Figure 4 This is a physical picture of the hardware of the embodiment of the present invention;

[0025] Figure 5 This is a multipath analysis diagram of pseudorange calculation according to an embodiment of the present invention;

[0026] Figure 6 is a comparison diagram of the measured trajectories of the positioning system according to an embodiment of the present invention;

[0027] Figure 7 It is a comparison diagram of the measured errors of the positioning system according to the embodiment of the present invention. DETAILED DESCRIPTION

[0028] The following is an explanation of the embodiments of the present invention by specific specific embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification, but the features of this invention are not limited to this embodiment. In order to provide a deep understanding of the present invention, many specific details will be included in the following description. The present invention can also be implemented without using these details. In addition, in order to avoid confusion or blurring the key points of the present invention, some specific details will be omitted in the description. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0029] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0030] like Figure 1 , a vehicle-mounted positioning system based on LTE signals, includes the following modules: a cell search module, a frame synchronization module, a pseudo-range calculation module, and an extended Kalman filter module.

[0031] Among them, the cell search module identifies base station signals near the area and distinguishes these base station signals;

[0032] The frame synchronization module uses PSS and SSS to complete the synchronization of LTE data frames and provide them to the pseudorange calculation module;

[0033] The pseudorange calculation module extracts pseudorange information from the LTE data frame. The present invention creatively uses the MDL criterion and the ESPRIT algorithm to extract the time delay from the base station to the receiver from the channel, thereby realizing pseudorange estimation.

[0034] The pseudorange estimation value obtained by the pseudorange calculation module is further processed by the extended Kalman filter module to reduce the error caused by the clock offset between the receiver and each base station, thereby outputting a smoother and more accurate position result.

[0035] The details are as follows:

[0036] 1.Cell search module

[0037] The input of the cell search module is the received baseband data, and the output is the cell ID of the base station signal existing in the signal frequency band demodulated from the baseband data. The cell search module can identify the base station signals near the area and distinguish these base station signals.

[0038] LTE signals from commercial base stations are received through USRP (Universal Software Radio Peripheral) devices. The antenna used can cover the LTE frequency band, that is, an omnidirectional antenna of 0-6GHz. The coverage range of commercial base stations is relatively long, usually reaching 1-3 kilometers, and the quality of the transmitted downlink signal is relatively high, which is conducive to the implementation of the entire positioning system.

[0039] In terms of hardware, the USRP device is controlled by a computer to complete the sampling of LTE signals. The center frequency is set to the frequency used by common commercial base stations, the default bandwidth is 20MHz, and the sampling rate is set to 30.72MHz. The entire sampling process is implemented through Labview, and the corresponding SDR module is used on Labview to build a receiver to complete the collection of LTE signals throughout the process.

[0040] The process of LTE cell search is as follows: LTE cell search is summarized as follows: the receiving end first performs rough synchronization on the received signal, that is, uses the locally generated ZC sequence to cross-correlate with the received signal, detects the PSS signal, completes the rough timing, and calculates Then, according to the determined PSS signal and Do fine synchronization, that is, determine the position of the SSS signal, complete frame synchronization and calculate Finally combined and It can be calculated The calculation formula is as follows:

[0041]

[0042] in, According to formula (1), it can be determined

[0043] 2. Frame synchronization module

[0044] LTE frame synchronization is the first important step to reconstruct LTE data frames, which is usually achieved through PSS and SSS in LTE signals. The input of this module is the demodulated baseband data, and the output is the synchronized tracked LTE data frame. This module determines the LTE data frame header by performing correlation detection on the PSS signal and the SSS signal, thereby tracking the LTE data frame.

[0045] The method of detecting PSS signal is relatively direct, and is also called the process of PSS blind detection. Based on the fact that PSS sequence is a ZC sequence, it has good orthogonality, which is well explained in the second part. The detection method is to generate three different ZC sequences locally according to three different root sequence numbers, and perform correlation detection with the received data. The detection method is as follows:

[0046]

[0047] In formula (2), r(n) is the received signal, s PSS (n) represents the PSS sequence generated locally by the receiver, n is the frame length, (.) * represents complex conjugate, (.) N represents a periodic shift operation, represents periodic convolution. Then take FFT and IFFT of formula (2), we can get:

[0048]

[0049] In formula (3), R(k) and Represent r(n) and s respectively PSS (n) The result after inverse fast Fourier transform. Similarly, after determining the position of PSS, coarse synchronization can be completed and the position of SSS can be calculated. The SSS signal can be used again to complete correlation detection to complete fine synchronization, and finally complete the synchronization of LTE data frames.

[0050] 3. Pseudorange calculation module

[0051] The input of the pseudorange calculation module is the synchronized LTE data frame, and the output is the pseudorange information extracted from the frame. This module uses the MDL criterion and ESPRIT algorithm for the CRS signal. The MDL principle is used to solve the source estimation problem and determine how many independent sources the data comes from. In this case, it is the multipath generated by the same base station signal reaching the receiver. The MDL principle is used to estimate the number of multipaths of the signal from the observed data. And according to the obtained multipath number Select the first Row construction matrix U s, get the signal subspace matrix, and then s Using the ESPRIT algorithm, the time delay information of each path for the signal to reach the receiver can be estimated, thereby extracting the time delay from the base station to the receiver in the channel and realizing pseudorange estimation.

[0052] The steps of this module are briefly summarized as follows:

[0053] (1) Construct the Hankel matrix X based on the channel frequency response CFR.

[0054] (2) Perform singular value decomposition on the data matrix X: X = UΣV H , and get the matrix U.

[0055] (3) Based on the dimensions K and P of the constructed data matrix X and the obtained singular value matrix Σ, the total description length MDL(γ) is constructed.

[0056] (4) Calculate γ when MDL(γ) takes the minimum value and obtain the estimated value of the multipath number

[0057] (5) Take from matrix U The largest eigenvalue is obtained to obtain the signal subspace matrix U s .

[0058] (6) Take U s The first (P-1) rows and the last (P-1) rows of s,1 and U s,2 .

[0059] (7) Based on the separated matrix U s,1 and U s,2 , use the least squares method (LS) to fit the transformation matrix Ψ

[0060] (8) Find the eigenvalues ​​of the matrix Ψ. After transformation, these eigenvalues ​​are the delay parameters to be estimated.

[0061] The specific process is as follows:

[0062] After completing LTE frame synchronization, the LTE data frame is mapped to RE (Resource Element), and the required pilot signal CRS is extracted from the corresponding RE according to the calculated cell ID. The estimated CFR (channel frequency response) is obtained using the least squares method. Construct the data matrix X as follows:

[0063]

[0064] The data matrix X is subjected to singular value decomposition, and the result is X = UΣV H, we get the value containing all singular values ​​σ1≥...≥σ P The diagonal matrix Σ of , formula (5) is the method for calculating MDL:

[0065]

[0066] In formula (5), where γ = 0...P-1, P = M-K+1, K is the smoothing coefficient when constructing the data matrix X, M is the number of subcarriers occupied by the CRS signal on an OFDM symbol, and P is the number of rows of the constructed data matrix. Therefore, the number of multipaths can be estimated as shown in formula (6):

[0067]

[0068] Get the number of multipaths Then, the signal subspace U is separated according to the unitary matrix U obtained by the singular value decomposition of the data matrix X. s， The ESPRIT algorithm is shown in formula (7):

[0069]

[0070] Finally, the rotation matrix Ψ is obtained and the eigenvalues ​​of the Ψ matrix are calculated Able to estimate The delay of each path is as follows:

[0071]

[0072] Formula (8) contains the delay information of all paths from the base station signal to the receiver, where the delay of the LOS path is The delay of the LOS path is the linear transmission time of the LTE signal directly from the commercial base station to the receiver, so the pseudo-range of the LOS path can be obtained. The pseudo-ranges from the vehicle to three different base stations can be obtained by three USRPs, and positioning can be achieved according to the TOA algorithm (existing technology).

[0073] 4. Extended Kalman filter module

[0074] The input of the extended Kalman filter module is the coordinate position of the receiver calculated from the pseudorange calculation module and the coordinate position of the known nearby base stations. The output is the estimated coordinate position of the receiver after Kalman filtering. This module uses Kalman filtering to reduce the error caused by the clock offset between the receiver and each base station, thereby outputting a smoother and more accurate position result.

[0075] In the estimation of pseudorange, the actual pseudorange estimation is formula (9):

[0076]

[0077] Where k is the time step, r r =[x r ,y r ] T is the position coordinate vector of the receiver, is the position coordinate vector of the u-th base station; δt r and are the clock deviation of the receiver and the clock deviation of the uth base station, respectively, u is the measurement noise, which has a mean of 0 and a variance of Based on these measurements, we discuss how to use EKF to estimate the state of the receiver.

[0078] First, define the state vector matrix X, including the position and velocity of the receiver, the clock deviation and clock drift between the receiver and each base station, as follows:

[0079]

[0080] in is the velocity vector of the receiver, Assume that the positioning system is sub The pseudo-range information is obtained once per second, so the system has a sampling period of T sub The discrete model for seconds is as follows:

[0081] x(k+1)=Fx(k)+w(k) (11)

[0083] Here w(k) is the process noise, and the state transfer matrix F is:

[0084]

[0085] At the same time, define the covariance matrix Q = diag[Q pv ,Q clk ],definition and is the spectral density of acceleration in the x and y directions, Q pv and Q clk The result is:

[0086]

[0087] By introducing the clock offset parameter into the covariance matrix and observing the clock offset as a state, this EKF design introduces the clock offset into the covariance matrix by putting the clock offset and clock drift into the state vector matrix for prediction, thereby reducing the error caused by the clock offset between the base station and the receiver, and further improving the positioning accuracy.

[0088] Figure 1 This is the structural diagram of the vehicle-mounted LTE positioning system. This vehicle-mounted positioning system focuses on solving the problem of poor GNSS positioning in complex GNSS environments. By receiving LTE signals, it first completes the cell search, determines the cell ID of the commercial base station, and completes the LTE frame synchronization. Then, the pseudo-range from the commercial base station to the receiver is estimated through the MDL criterion and ESPRIT algorithm, and the positioning is achieved through the TOA method. Finally, the extended Kalman filter is introduced to eliminate the clock bias and improve the positioning accuracy.

[0089] Figure 2 This is the structural diagram of the LTE signal cell search module. Cell search is the first step in achieving frame synchronization for LTE signals. Baseband data is used to perform PSS correlation detection to achieve coarse synchronization and SSS correlation detection to achieve fine synchronization, and finally the cell ID is solved to determine which base station the received signal comes from, achieve frame synchronization, and reconstruct the LTE data frame.

[0090] Figure 3 This is the structural diagram of the key positioning algorithm. According to the CRS signal obtained in the LTE data frame, the least squares method is used to estimate the channel to obtain the CFR, and then the MDL criterion is used to estimate the number of multipaths and the ESPRIT algorithm is used to estimate the pseudorange. Finally, the EKF is used to complete the positioning.

[0091] Figure 4 The following is a physical picture of the hardware of this embodiment. The USRP device model used is B210, and the three receiving antennas are bipolar omnidirectional antennas with a frequency band of 0-6GHz. The clocks of the three USRPs are timed and synchronized through a high-precision clock GPSDO, and the receiving signal control of the three USRPs is completed on a computer.

[0092] Figure 5 This is a simulation test diagram. According to the MDL criterion and ESPRIT algorithm proposed in the present invention, Matlab is used for simulation testing. It can be seen from the image that there are three paths from the transmitter to the receiver, and the multipath distinction effect is relatively good. The delay of the three paths can be distinguished in most test data, which verifies the robustness and practicality of this algorithm.

[0093] Figure 6This is a comparison chart of the measured trajectory of the positioning system. It can be seen that the trajectory after EKF filtering is still relatively close to the real trajectory. Therefore, in the scenario where the GNSS signal is lost, it is completely possible to rely on the LTE signal for positioning, which meets the requirements in terms of accuracy.

[0094] Figure 7 It is a curve of the RMSE of the measured longitude and latitude error of the positioning system changing with time. The horizontal axis is time, the unit is data frame, and the vertical axis is RMSE error, the unit is meter. It can be seen that the fluctuation of the whole system is not large after EKF, and the error is basically maintained at a relatively stable level. Finally, the measured longitude RMSE of the whole system is 8.4661m, and the latitude error is 7.7741m, which is close to the GNSS positioning accuracy.

[0095] Although the present invention has been illustrated and described with reference to certain preferred embodiments of the present invention, it should be understood by those skilled in the art that the above is a further detailed description of the present invention in conjunction with specific embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. Those skilled in the art may make various changes in form and details, including making several simple deductions or substitutions, without departing from the spirit and scope of the present invention.

Claims

1. A vehicle positioning system based on LTE signals, characterized in that: It includes the following modules: cell search module, frame synchronization module, pseudo-range calculation module, and extended Kalman filter module; Among them, the cell search module identifies base station signals near the area and distinguishes these base station signals; The frame synchronization module uses PSS and SSS to complete the synchronization of LTE data frames and provide them to the pseudorange calculation module; The pseudorange calculation module extracts pseudorange information from the LTE data frame and uses the MDL criterion and ESPRIT algorithm to extract the time delay from the base station to the receiver from the channel, thereby realizing pseudorange estimation; The pseudorange estimation value obtained by the pseudorange calculation module is further processed by the extended Kalman filter module to reduce the error caused by the clock offset between the receiver and each base station, thereby outputting a smoother and more accurate position result.

2. The vehicle-mounted positioning system based on LTE signals as claimed in claim 1, characterized in that: The cell search module receives as input the received baseband data and outputs the cell ID of the base station signal existing in the signal frequency band demodulated from the baseband data. The cell search module identifies the base station signals near the area and distinguishes these base station signals; The cell search module receives LTE signals from commercial base stations through USRP devices, and the antenna uses an omnidirectional antenna that can cover the LTE frequency band, that is, 0-6GHz.

3. The vehicle-mounted positioning system based on LTE signals as claimed in claim 1, characterized in that: The input of the pseudorange calculation module is the synchronized LTE data frame, and the output is the pseudorange information extracted from the frame. The CRS signal is solved by using the MDL principle and the ESPRIT algorithm: The multipath number of the signal is estimated from the observed data using the MDL principle. And according to the obtained multipath number Select the first Row construction matrix U s , get the signal subspace matrix, and then s The ESPRIT algorithm is used to estimate the time delay information of each path from the signal to the receiver, thereby extracting the time delay from the base station to the receiver in the channel and realizing pseudorange estimation.

4. The vehicle-mounted positioning system based on LTE signals as claimed in claim 1, characterized in that: The pseudorange calculation module processes as follows: (1) Construct the Hankel matrix X according to the channel frequency response CFR; (2) Perform singular value decomposition on the data matrix X: X = UΣV H , get the matrix U; (3) Construct the total description length MDL(γ) based on the constructed data matrix X dimensions K, P, and the obtained singular value matrix Σ; (4) Calculate γ when MDL(γ) takes the minimum value and obtain the estimated value of the multipath number (5) Take from matrix U The largest eigenvalue is obtained, and the signal subspace matrix U s ; (6) Take U s The first (P-1) rows and the last (P-1) rows of s,1 and U s,2 ; (7) Based on the separated matrix U s,1 and U s,2 , use the least squares method to fit the transformation matrix Ψ; (8) Find the eigenvalues ​​of the matrix Ψ. After transformation, these eigenvalues ​​are the delay parameters to be estimated.

Images