TDOA positioning method based on data compensation

Abstract

The application provides a TDOA positioning method based on data compensation, solves the problems of loss of position of a non-cooperative radiation source at the moment and decrease of utilization rate of measured data caused by non-cooperative reception of the signal of the sensor node to the non-cooperative radiation source, and the steps are as follows: time difference data is acquired to generate a time difference matrix; then, whether the radiation source at a historical moment satisfies a common view condition is judged; for the moment which does not satisfy the common view condition, the time difference of arrival (TDOA) values of the same pair of nodes at continuous multiple moments nearby are polynomially fitted, the TDOA values after the polynomial fitting are used to supplement the lost TDOA values, full-rank TDOA measurement equations are constructed, and the position of the radiation source at the non-common view moment is calculated. The application improves the utilization rate of the TDOA data of the non-cooperative radiation source and improves the positioning rate of the non-cooperative radiation source.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of communication technology, and further relates to a data compensation-based TDOA (Time Difference of Arrival) positioning method. The present application can be used for non-cooperative radiators that emit continuous signals, and in particular for a data compensation-based TDOA positioning method. BACKGROUND

[0002] With the rapid development of mobile communication technology, electromagnetic spectrum signal source positioning technology has become a necessary function for the next generation of mobile communication systems. In recent years, research results have shown that, since electromagnetic spectrum monitoring work is mostly passive monitoring of signals, it is impossible to control the monitored targets, so strict clock synchronization cannot be achieved between the monitoring targets and the monitoring nodes. The TDOA positioning method only requires clock synchronization between the monitoring nodes, and can be applied to various types of networks such as cellular networks and wireless sensor networks. Due to its low application cost and high positioning accuracy, the positioning method has attracted widespread attention and has been determined as a standard positioning method in 3GPP.

[0003] The TDOA positioning method consists of two steps: 1. estimating the propagation time delay difference of radio signals in the air; 2. calculating the position of the target node using the time delay difference. The time delay difference, simply referred to as the time delay, is the time difference between the same source signals received by different monitoring nodes due to the different propagation distances of the monitored target emitted radio signals to different monitoring nodes. Time delay difference estimation refers to using the theory and method of parameter estimation and signal processing to accurately and quickly estimate the time delay difference between different monitoring nodes due to different monitoring signal propagation paths according to the received monitoring signals. From this, other related target parameters such as the distance, azimuth, direction of motion and speed of the target node are further determined. Network positioning is to calculate the distance difference between the target node and the monitoring node using the time delay difference, so accurate and fast estimation of the time delay difference is the premise of network positioning.

[0004] Time delay difference estimation technology is a new technology developed in the past ten years, and its application in practical engineering has attracted great attention. Abroad, time delay difference estimation technology has been widely used in the fields of military, geophysics, biomedical and industrial processes. In China, time delay difference estimation technology has also begun to be applied in the field of passive positioning and tracking. There are various time delay difference estimation algorithms available at present, such as basic correlation method, generalized correlation function GCC method, generalized phase spectrum method, adaptive LMS filter method, etc.

[0005] In practical applications, due to the target non-cooperation, obstacle shielding and other reasons, all signals from the radiation source cannot be received, resulting in signal loss. The traditional observation requires that multiple sensor nodes can receive the same signal, and the loss of signal will cause the traditional observation to decrease, thereby affecting the positioning accuracy or failing to position. Due to the non-common view of some sensor nodes, the signal of the non-cooperative radiation source cannot be received, resulting in the loss of the position of the non-cooperative radiation source at that moment and the decrease of the utilization rate of the measurement data. SUMMARY

[0006] The purpose of the present application is to overcome the shortcomings of the prior art, and to provide a TDOA positioning method based on data compensation, which aims to solve the problem of loss of non-cooperative radiation source position at that moment and decrease of measurement data utilization rate due to non-common view of some sensor nodes.

[0007] The idea to achieve the above purpose is that the present application performs polynomial fitting on the time difference of arrival TDOA values of the same pair of nodes at continuous multiple moments near the moment that does not meet the common view condition, uses the TDOA values after polynomial fitting to supplement the TDOA values of the lost moment, and constructs a full-rank measurement equation with the supplemented TDOA values and the actually measured TDOA values that cannot be utilized, to calculate the position of the non-cooperative radiation source at the lost TDOA moment. This idea compensates for the lost TDOA values of the non-cooperative radiation source, thereby calculating the position of the lost non-cooperative radiation source. The supplemented TDOA values can be used to construct a full-rank measurement equation with the actually measured TDOA values that cannot be utilized, thereby improving the utilization rate of the measurement data.

[0008] The technical scheme of the present application comprises the following steps:

[0009] Step 1, obtain time difference data:

[0010] Step 1.1, select one node in M sensor nodes as the master station, and the other nodes as the auxiliary stations, wherein M≥4;

[0011] Step 1.2, calculate the time difference of arrival TDOA value of the electromagnetic wave of the target radiation to be positioned to the master station and each auxiliary station using the time difference of arrival formula; if the TDOA value cannot be calculated due to the loss of the signal of the radiation source arriving at the master station or the auxiliary station at a certain moment, an abnormal TDOA value 10s that cannot be obtained is used to represent it;

[0012] Step 1.3, generate a time difference matrix with all the time difference of arrival TDOA values, the number of rows of the matrix represents the total number of sensors-1, the number of columns represents the total number of moments, and each element value represents the TDOA value measured by the corresponding sensor node pair at the corresponding moment;

[0013] Step 2, generating a non-locatable radiation source matrix:

[0014] Taking the time in the time difference matrix satisfying the common view condition as a locatable radiation source matrix, and taking the time not satisfying the common view condition as a non-locatable radiation source matrix;

[0015] Step 3, fitting TDOA values:

[0016] Step 3.1, performing polynomial fitting on the time difference TDOA values of the same pair of nodes for the time not satisfying the common view condition and the continuous multiple times near the time;

[0017] Step 3.2, inputting each time value in the row of the time difference matrix except the abnormal value and the corresponding TDOA value of the time into the matlab polynomial fitting function to obtain the fitted TDOA value;

[0018] Step 4, calculating the compensation value of the TDOA abnormal value of each time by using the following formula:

[0019]

[0020] Wherein, TDOA k represents the TDOA compensation value of the kth time, TDOA k-1 represents the TDOA value of the k-1th time, TDOA n represents the TDOA value of the nth time, n>k;

[0021] Step 5, calculating the position of the radiation source at the non-common view time:

[0022] Inputting the TDOA compensation value and the TDOA value in the non-locatable radiation source matrix into the full rank measurement equation to calculate the position of the radiation source.

[0023] Compared with the prior art, the present application has the following advantages:

[0024] Firstly, since the present application performs polynomial fitting on the time difference TDOA values of the same pair of nodes for the time not satisfying the common view condition and the continuous multiple times near the time, the lost TDOA value of the time is supplemented by using the fitted time difference TDOA value, which overcomes the deficiency that the loss of the non-cooperative radiation source TDOA parameter in the prior art leads to the loss of the position of the non-cooperative radiation source at the time, so that the present application improves the positioning rate of the non-cooperative radiation source.

[0025] Secondly, the TDOA positioning parameters supplemented by the application can construct full-rank measurement equations with the actual measured TDOA values that cannot be utilized, so as to overcome the deficiency of the decline of the utilization rate of measurement data caused by the loss of non-cooperative radiation source TDOA parameters in the prior art, so that the application improves the TDOA utilization rate of the non-cooperative radiation source, so that the position of the non-cooperative radiation source at the lost moment is more accurate. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is a flowchart of the application;

[0027] Figure 2 is a target real trajectory, original positioning result and result after supplementing the lost point when the TDOA parameter loss rate of the application is 0.2;

[0028] Figure 3 is a trend chart of the non-compensated trajectory positioning rate, the trajectory compensation positioning rate at the parameter moment and the trajectory compensation positioning rate without the parameter moment under different TDOA parameter loss rates of the application;

[0029] Figure 4 is a result chart of the RMSE of the trajectory after supplementing the lost point and the real trajectory and the RMSE of the interrupted fixed positioning result and the real trajectory when the TDOA parameter loss rate of the application is 0.2. DETAILED DESCRIPTION

[0030] The application will be further described below in combination with the drawings and examples.

[0031] Reference Figure 1 The implementation steps of the embodiment of the application will be further described.

[0032] Step 1, acquiring time difference data

[0033] Step 1.1, selecting one node in the M sensor nodes as a master station and the other nodes as auxiliary stations, wherein M is greater than or equal to 4;

[0034] Step 1.2, calculating the time difference of arrival (TDOA) value of the electromagnetic wave of the target radiation to be positioned at different moments by using the time difference of arrival formula; if the TDOA value cannot be calculated due to the loss of the signal of the radiation source signal arriving at the master station or the auxiliary station at a certain moment, an abnormal TDOA value 10s that cannot be obtained is used to represent it;

[0035] Step 1.3, generating a time difference matrix of all time differences of arrival (TDOA) values, wherein the number of rows represents the total number of sensor nodes minus 1, the number of columns represents the total number of moments, and each element value represents the TDOA value measured by the corresponding sensor node pair at the corresponding moment;

[0036] The time difference of arrival formula is as follows:

[0037]

[0038]

[0039] Wherein, t i1 represents the time difference TDOA value of electromagnetic waves of the target radiation to be positioned reaching the main station and the i th auxiliary station, i = 2,..., M, r i1 represents the distance difference between the non-cooperative radiation source reaching the i th auxiliary station and the main station, c represents the propagation speed of light, ||·|| represents the Euclidean norm operation, (·) 0 represents the true value of the content in the parentheses, u represents the position of the non-cooperative radiation source, s i represents the position of the i th auxiliary station, n represents the zero-mean Gaussian white noise of the distance difference between the i th auxiliary station and the main station.

[0040] The specific operation steps are as follows: first, four sensor nodes in three-dimensional space are selected, the first node is taken as the main station, and the other nodes are taken as auxiliary stations. Then, the time difference of arrival formula described in step 1.2 is used to calculate t 21 , t 31 , t 41 , and the uncalculated TDOA is represented by TDOA abnormal value 10s, and finally a time difference matrix is generated:

[0041]

[0042] t i1 represents the time difference TDOA value of electromagnetic waves of the target radiation to be positioned reaching the main station and the i th auxiliary station, i = 2,..., 4, t i1 k represents the t i1 value at the k th time, k = 1,..., 6.

[0043] Step 2, the time difference matrix is used as the locatable radiation source matrix which meets the common view condition; the time difference matrix which does not meet the common view condition is used as the unlocatable radiation source matrix;

[0044] The common view condition is that the positioning method based on TDOA parameters requires that multiple observation stations must receive the signal of the radiation source at the same time. In two-dimensional space, two or more TDOAs can determine the position of the radiation source, and in three-dimensional space, three or more TDOAs can determine the position of the radiation source; the locatable radiation source matrix stores TDOA normal values; the unlocatable radiation source matrix stores TDOA normal values and abnormal values.

[0045] The specific operation steps are as follows: observing the time difference matrix, finding that there are TDOA abnormal values at the 2nd and 5th moments, causing the radiation sources at the 2nd and 5th moments to not satisfy the common view condition, and obtaining the matrix of locatable radiation sources and unlocatable radiation sources:

[0046]

[0047] Step 3, polynomial fitting is performed on the TDOA values.

[0048] Each time value in the row of the time difference matrix except the abnormal value is input into the matlab polynomial fitting function together with the TDOA value at the corresponding moment, and the fitted TDOA value is obtained;

[0049] The polynomial fitting is fitting all observation points in a small analysis region containing a plurality of analysis grid points with a polynomial expansion

[0050] The mathematical expression is:

[0051]

[0052] Wherein, M is the highest degree of the polynomial, the value of M is less than the total number of input samples, x represents the time value, ω j represents the coefficient of the polynomial, and y represents the TDOA value at the corresponding moment obtained after inputting the time value.

[0053] The specific operation steps are as follows:

[0054] 1, 2, 3, 4, 6 and t 21 1, t 21 2, t 21 3, t 21 4, t 21 6 are input into the matlab polynomial fitting function to obtain the fitted TDOA value: 1, 3, 4, 5, 6 and t 31 1, t 31 3, t 31 4, t 31 5, t 31 6 are input into the matlab polynomial fitting function to obtain the fitted TDOA value:

[0055] Step 4, the compensation value of the TDOA abnormal value at each moment is calculated by the following formula:

[0056]

[0057] Wherein, TDOA kTDOA compensation value at the kth moment, TDOA k-1 TDOA value at the k-1th moment, TDOA n TDOA value at the nth moment, n>k.

[0058] The specific operation steps are as follows: because there are abnormal values at the 2nd and 5th moments, the compensation value calculation results are as follows:

[0059]

[0060]

[0061] Step 5, calculate the position of the radiation source under the non-common view condition.

[0062] The TDOA compensation value and the TDOA value in the unlocatable radiation source matrix are input into the full rank measurement equation to calculate the position of the radiation source.

[0063] Because two or more TDOAs can determine the position of the radiation source in a two-dimensional space, and three or more TDOAs can determine the position of the radiation source in a three-dimensional space. The full rank measurement equation refers to the linear independence of the measurement equation coefficient and the solvability of the position of the radiation source. The classical positioning calculation method based on TDOA is divided into two categories, one is the algorithm that can obtain an analytical solution, such as Fang algorithm and Chan algorithm, and the other is the iterative algorithm, such as Taylor algorithm; here we adopt Chan algorithm.

[0064] The effect of the present application will be further described below in combination with a simulation experiment.

[0065] 1. Simulation experiment conditions:

[0066] The hardware platform of the simulation experiment of the present application is: the processor is AMD Ryzen 75800H with Radeon Graphics, the main frequency is 3.20GHz, and the memory is 16.0GB.

[0067] The software platform of the simulation experiment of the present application is: Windows 11 operating system and Matlab R2021b.

[0068] The simulation experiment scenario of this invention consists of a single maneuvering aircraft target and four sensor nodes. The single maneuvering aircraft target moves at a constant linear velocity of 100 m / s from the start of flight to the 10th moment. From the 11th to the 15th moment, it moves in a sinusoidal motion with an amplitude of 200 m. From the 16th to the 30th moment, it moves at a constant linear velocity of 100 m / s. Each moment contains a position error of size randn(1)*sqrt(R), where R = 10^2 m. The four sensor nodes are respectively set at coordinates [0, 0], [1000, 250], [5000, 0], and [5000, 5000] in a two-dimensional Cartesian coordinate system.

[0069] 2. Simulation content and result analysis:

[0070] This invention uses a sensor node located at [0, 0] as the reference node in the simulation experiment. It calculates the arrival time difference (TDOA) between the signal of a single maneuvering aircraft target at each moment and the other three sensor nodes and the reference node. All arrival time differences are grouped into a 3*30 matrix. The number of rows in the matrix is ​​the total number of sensors – 1, and the number of columns is the total number of moments. Each column stores the TDOA at the current moment, including a TDOA measurement error of 50 ns. The Chan algorithm is then used to calculate the trajectory of the single maneuvering aircraft target. A loss rate of 0.2 is set for the first row of the TDOA matrix. Due to the loss of TDOA values, the TDOA matrix is ​​not full rank, thus failing to provide a positioning result. This results in the inability to locate the position of a single maneuvering aircraft target at some moments. The Chan algorithm is then used to calculate the trajectory of the moments where the position of a single maneuvering aircraft target is lost. The method of this invention is used to compensate for the moments with lost TDOA values. The compensated TDOA is then reconstructed with the TDOA measurement value at that moment, and the Chan positioning algorithm is used to obtain the estimated value of the target at that moment.

[0071] Figure 2 This is a simulation result diagram when the TDOA loss rate is 0.2. Figure 2 The curves marked with squares represent the actual trajectory curves; the curves marked with triangles represent the original positioning result curves; and the curves marked with circles represent the trajectory curves after the missing points are supplemented. Figure 2The horizontal and vertical axes in the diagram represent the horizontal and vertical coordinates of a single aircraft target in a Cartesian coordinate system, respectively. Without setting a loss rate for the TDOA matrix, the Chan algorithm is used to calculate the true trajectory of a single aircraft target. With a loss rate of 0.2 set for the TDOA matrix, the Chan algorithm is used again to calculate the original positioning result of the single aircraft target. The method of this invention compensates for moments with lost TDOA data. The compensated TDOA is then reconstructed with the TDOA measurement value at that moment, and the Chan positioning algorithm is used to obtain the estimated value of the target at that moment, resulting in the trajectory after the lost points are replenished.

[0072] Figure 3 The figure shows the simulation results obtained after 2000 Monte Carlo cycles when the TDOA loss rate gradually increased from 0.1 to 0.5. Figure 3 The curves marked with squares represent the positioning rate without compensation. The curves marked with circles represent the positioning rate with trajectory compensation at parameterized times. The curves marked with triangles represent the positioning rate with compensated trajectory at parameterless times. Figure 3 The horizontal and vertical axes represent the loss rate of TDOA and the localization rate of a single aircraft maneuvering target, respectively. The graph shows the uncompensated trajectory localization rate curves obtained when different loss rates are set for the TDOA matrix. By setting different loss rates for the TDOA matrix, but only for moments with TDOA parameters, the method of this invention is used for compensation to obtain the trajectory-compensated localization rate curve for moments with parameters. Furthermore, by setting different loss rates for the TDOA matrix, but using the method of this invention for both moments with and without TDOA parameters, the method of this invention is applied to obtain the compensated trajectory localization rate curve for moments without parameters.

[0073] Figure 4 This is a scatter plot of the root mean square error (RMSE) with a loss rate of 0.2 after 2000 Monte Carlo cycles. Figure 4 The scattered points marked with squares represent the RMSE after supplementing the lost points using the method of this invention, while the scattered points marked with circles represent the RMSE after interrupting the fixed positioning results. Figure 4 The horizontal and vertical axes in the figure represent time and root mean square error, respectively.

[0074] The following is combined Figure 2 , Figure 3 , Figure 4 The simulation diagrams further illustrate the effects of the present invention.

[0075] Depend on Figure 2 It can be concluded that, under the condition of a TDOA loss rate of 0.2, the trajectory after supplementing the lost points has a higher degree of fit with the true trajectory compared with the original positioning result.

[0076] Depend onFigure 3 It can be concluded that the positioning rate of trajectory compensation at the parameter moment is greatly improved compared with the non-compensation trajectory positioning rate, wherein the positioning rate is improved by about 0.15 when the loss rate is 0.2, and by about 0.22 when the loss rate is 0.5.

[0077] By Figure 4 It can be concluded that the RMSE after supplementing the loss point is about 40 m lower than the RMSE after interrupting the fixed positioning result, which shows that the supplemented loss point has good positioning accuracy.

Claims

1. A data compensation based TDOA positioning method, characterized in that, The time difference of arrival (TDOA) values of the same pair of nodes at continuous multiple time points near the time point that does not satisfy the common view condition are polynomial fitted, and the time difference of arrival (TDOA) values after the polynomial fitting are used to supplement the lost TDOA values; the steps of the positioning method include the following: Step 1, obtaining time difference data: Step 1.1, selecting an optional node in M sensor nodes as a master station, and other nodes as auxiliary stations, wherein M≥4; Step 1.2, calculating the time difference of arrival (TDOA) values of the electromagnetic waves radiated by the target to be positioned to the master station and each auxiliary station by using the time difference of arrival (TDOA) formula; if the TDOA value cannot be calculated due to the loss of the signal of the radiation source signal arriving at the master station or the auxiliary station at a certain time, an impossible TDOA abnormal value 10s is used to represent it; Step 1.3, generating a time difference matrix from all the time difference of arrival (TDOA) values, wherein the number of rows represents the total number of sensors-1, the number of columns represents the total number of time points, and each element value represents the TDOA value measured by the corresponding sensor node pair at the corresponding time point; Step 2, generating an unlocatable radiation source matrix: The time points in the time difference matrix that satisfy the common view condition are used as the locatable radiation source matrix; the time points that do not satisfy the common view condition are used as the unlocatable radiation source matrix; Step 3, fitting the TDOA value: Step 3.1, polynomial fitting the time difference of arrival (TDOA) values of the same pair of nodes at continuous multiple time points near the time point that does not satisfy the common view condition; Step 3.2, inputting each time point value in the row of the abnormal value in the time difference matrix except the abnormal value and the corresponding time point TDOA value into the matlab polynomial fitting function to obtain the fitted TDOA value; Step 4, calculating the compensation value of the TDOA abnormal value at each time point by using the following formula: wherein TDOA k represents the TDOA value at the kth time point, TDOA k-1 represents the TDOA value at the k-1th time point, TDOA n represents the TDOA value at the nth time point, n>k; Step 5, calculating the position of the radiation source at the non-common view time point: Inputting the TDOA compensation value and the TDOA value in the unlocatable radiation source matrix into the full-rank measurement equation to calculate the position of the radiation source.

2. The data compensation based TDOA positioning method according to claim 1, characterized in that: The time difference of arrival (TDOA) formula in step 1.2 is as follows: where t i1 represents the time difference of arrival (TDOA) value of the electromagnetic wave of the target radiation to be positioned arriving at the main station and the i-th auxiliary station, i = 2,...,M, r i1 represents the distance difference between the non-cooperative radiation source arriving at the i-th auxiliary station and the main station, c represents the propagation speed of light, ||·|| represents the Euclidean norm operation, (·) 0 represents the true value of the content in the parentheses, u represents the position of the non-cooperative radiation source, s i represents the position of the i-th auxiliary station, n i1 represents the zero-mean Gaussian white noise of the distance difference between the i-th auxiliary station and the main station.

3. The data compensation based TDOA positioning method according to claim 1, characterized in that, The common view condition in step 2 refers to that all sensor nodes simultaneously receive the signal of the radiation source at the current time point.

Images