Signal tracking method, modem and time synchronization system

Abstract

The application discloses a signal tracking method, a modem and a time synchronization system, and relates to the technical field of high-precision time synchronization. The method comprises the following steps: acquiring a time comparison signal; generating a first measurement value of a carrier parameter by using updated carrier information of a previous tracking loop and the time comparison signal; calculating a measurement noise covariance matrix according to a first iteration parameter value of a strong tracking algorithm in the previous tracking loop, a first estimated value of the carrier parameter in the previous tracking loop and the first measurement value; calculating a second estimated value of the carrier parameter in a current tracking loop by using the measurement noise covariance matrix and a second iteration parameter value of the strong tracking algorithm in the current tracking loop; and updating the carrier information by using the second estimated value to obtain a signal tracking result of the current tracking loop. The application can improve the accuracy of the signal tracking result of the tracking loop, and effectively improve the robustness and accuracy of the tracking loop in the time comparison modem in a high dynamic environment.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of high-precision time synchronization, and in particular to a signal tracking method, a modem and a time synchronization system. BACKGROUND

[0002] With the development of science and technology, the precision of time is required higher and higher in various industries, and time service has become an important cornerstone for the development of a country. High-precision time synchronization is the key to realizing time service, and has very important application value in the fields of time measurement, navigation and positioning, space measurement and control, communication radar and the like. At present, in unmanned aerial vehicle formation, vehicle-mounted and the like, high-precision time synchronization is a basic guarantee for the normal work of the system. However, in these systems, unmanned aerial vehicles, vehicles and the like are often in a high dynamic scene, and therefore, the time synchronization system also needs to work in a high dynamic platform. When the time synchronization system is in a high dynamic state, the time synchronization precision is greatly reduced. At present, time synchronization mainly works in a static scene, and there are still very few time synchronization systems designed for high dynamic scenes.

[0003] In order to improve the performance of the time synchronization system in the high dynamic platform such as airborne and vehicle-mounted, it is urgent to improve the performance of the time synchronization system in the high dynamic scene. Time synchronization technology includes one-way time transfer technology and two-way time transfer technology, and the two-way time transfer technology is mainly used for high-precision time synchronization at present because of its symmetric propagation path and high synchronization precision. In the two-way time synchronization system, the modem is a core component and directly determines the time synchronization precision. The tracking loop is an important part of the modem, and the tracking loop is easy to lose lock in a high dynamic scene. In view of the above problems, no effective solution has been proposed. SUMMARY

[0004] Embodiments of the present application provide a signal tracking method, a modem and a time synchronization system to at least solve the technical problems of poor robustness and low tracking precision of the tracking loop in a high dynamic scene.

[0005] According to an aspect of the embodiments of the present application, a signal tracking method is provided, including: acquiring a time comparison signal; generating a first measurement value of a carrier parameter by using updated carrier information of a last tracking loop and the time comparison signal; calculating a measurement noise covariance matrix according to a first iteration parameter value of a strong tracking algorithm in the last tracking loop, a first estimated value of the carrier parameter in the last tracking loop and the first measurement value; calculating a second estimated value of the carrier parameter in a current tracking loop by using the measurement noise covariance matrix and a second iteration parameter value of the strong tracking algorithm in the current tracking loop; and updating the carrier information by using the second estimated value to obtain a signal tracking result of the current tracking loop.

[0006] Optionally, the carrier parameter comprises at least a carrier phase error parameter, a carrier Doppler shift parameter and a carrier Doppler rate parameter.

[0007] Optionally, the first measurement value of the carrier parameter is generated by using the updated carrier information of the last tracking loop and the time comparison signal, comprising: generating pseudo code information; demodulating the time comparison signal by using the updated carrier information of the last tracking loop and the pseudo code information to obtain a demodulation result; filtering the target frequency information in the demodulation result to obtain an integration result; calculating the carrier phase error by using the integration result to obtain the first measurement value of the carrier parameter.

[0008] Optionally, the measurement noise covariance matrix is calculated according to the first iteration parameter value of the strong tracking algorithm in the last tracking loop, the first estimation value of the carrier parameter in the last tracking loop and the first measurement value, comprising: obtaining the first degree of freedom number and the first inverse scale matrix of the measurement noise covariance matrix in the last variational Bayes iteration; calculating the second degree of freedom number of the measurement noise covariance matrix in the current variational Bayes iteration by using the first degree of freedom number; calculating the second inverse scale matrix of the measurement noise covariance matrix in the current variational Bayes iteration by using the first inverse scale matrix, the first iteration parameter value, the first estimation value and the first measurement value; calculating the measurement noise covariance matrix by using the second degree of freedom number, the second inverse scale matrix and the dimension value of the first measurement value.

[0009] Optionally, the measurement noise covariance matrix is calculated according to the following formula:

[0010]

[0011]

[0012]

[0013]

[0014] wherein, denotes the degree of freedom number, denotes the inverse scale matrix, i denotes the number of variational iterations, k denotes the number of STKF iterations, z k denotes the measurement vector at time k, H represents the measurement matrix, denotes the state variable at time k, P k is the estimated state covariance matrix, m denotes the dimension of the measurement vector, denotes the measurement noise covariance matrix.

[0015] Optionally, the second estimated value of the carrier parameter in the current tracking loop is calculated by using the measurement noise covariance matrix and the second iteration parameter value of the strong tracking algorithm in the current tracking loop, comprising: calculating a fading factor by using the measurement noise covariance matrix; calculating a predicted state covariance matrix by using the fading factor and the second iteration parameter value; and calculating the second estimated value of the carrier parameter in the current tracking loop by using the predicted state covariance matrix.

[0016] Optionally, the fading factor is calculated according to the following formula:

[0017] λ k = diag(λ 1,k ,λ 2,k …,λ n,k )

[0018]

[0019]

[0020]

[0021]

[0022]

[0023] wherein λ k represents the fading factor, σ i is a predefined constant, tr[·] represents the trace of a matrix, V k represents the estimated innovation covariance matrix, R k represents the measurement noise covariance matrix, H represents the measurement matrix, Q k-1 represents the system noise covariance, Φ represents the system transition matrix, represents the estimated state covariance matrix, and d represents the innovation sequence.

[0024] Optionally, the predicted state covariance matrix is calculated according to the following formula:

[0025]

[0026] wherein P represents the predicted state covariance matrix, λ k represents the fading factor, Φ represents the system transition matrix, represents the estimated state covariance matrix, and Q k-1 represents the system noise covariance.

[0027] According to another aspect of the embodiments of the present application, a modem is also provided, comprising: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to perform the above-mentioned signal tracking method.

[0028] According to another aspect of the embodiments of the present application, a time synchronization system is also provided, comprising the above-mentioned modem.

[0029] In the embodiments of the present application, a time comparison signal is acquired; a first measurement value of a carrier parameter is generated by using updated carrier information of a last tracking loop and the time comparison signal; a measurement noise covariance matrix is calculated according to a first iteration parameter value of a strong tracking algorithm in the last tracking loop, a first estimated value of the carrier parameter in the last tracking loop and the first measurement value; a second estimated value of the carrier parameter in a current tracking loop is calculated by using the measurement noise covariance matrix and a second iteration parameter value of the strong tracking algorithm in the current tracking loop; and the carrier information is updated by using the second estimated value to obtain a signal tracking result of the current tracking loop. In the embodiments of the present application, the measurement noise covariance matrix can be adaptively adjusted in each tracking loop, and the second estimated value of the carrier parameter in the strong tracking algorithm of the tracking loop can be calculated based on the adjusted measurement noise covariance matrix, so that the accuracy of the second estimated value of the carrier parameter in the strong tracking algorithm of the tracking loop can be improved, the accuracy of the signal tracking result of the tracking loop can be improved, and the robustness and accuracy of the tracking loop in the time comparison modem in a high dynamic environment can be effectively improved. BRIEF DESCRIPTION OF DRAWINGS

[0030] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:

[0031] Figure 1 A flow chart of the signal tracking method provided by the embodiments of the present application;

[0032] Figure 2 A block diagram of the STKF tracking loop based on the variational Bayesian provided by the optional embodiments of the present application;

[0033] Figure 3 An algorithm flow chart of the STKF tracking loop based on the variational Bayesian provided by the optional embodiments of the present application;

[0034] Figure 4 A schematic diagram of the signal tracking device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0035] Embodiments of the present application will be described below in more detail with reference to the accompanying drawings. While certain embodiments of the present application are shown in the drawings, it is understood that the present application can be embodied in various forms and should not be interpreted in a limited sense as set forth in the embodiments set forth herein, but rather the embodiments are provided to more thoroughly and completely understand the present application. It should be understood that the drawings and embodiments of the present application are for illustrative purposes only and are not intended to limit the scope of protection of the present application.

[0036] It should be understood that each step described in the method embodiments of the present application can be performed in different orders and / or in parallel. In addition, the method embodiments can include additional steps and / or omit the steps shown. The scope of the present application is not limited in this respect.

[0037] The term "comprising" and variations thereof as used herein are open-ended, that is "including, but not limited to". The term "based on" means "based, at least in part, on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Related terms are defined as follows. It should be noted that reference to a "first", "second", etc. concept in the present application is merely used to differentiate different devices, modules or units, and does not imply a sequence or interdependence of the functions performed by these devices, modules or units.

[0038] It should be noted that the modification of "one", "multiple" mentioned in the present application is illustrative and not restrictive, and those skilled in the art should understand that unless otherwise explicitly indicated in the context, it should be understood as "one or more".

[0039] The names of the messages or information exchanged between the devices in the embodiments of the present application are only for illustrative purposes, and are not intended to limit the scope of the messages or information.

[0040] The existing time comparison tracking loop uses a phase-locked loop, which is easy to lose lock directly in a high dynamic scene, causing the time synchronization system to fail to work normally and failing to meet the time synchronization requirements of gradually developing airborne, vehicle-mounted and other high dynamic platforms. In order to improve the performance of the time synchronization system, it is first necessary to improve the performance of the tracking loop under high dynamic platforms.

[0041] Considering that the measurement noise covariance R is fixed in the traditional STKF (Strong Tracking Kalman Filter) tracking, but from the process of calculating the fading factor, it can be known that the inaccurate measurement noise covariance R will affect the accuracy of the fading factor, thereby causing the inaccurate filtering result of the whole tracking loop. In a high dynamic environment, the measurement noise is constantly changing, and the fixed R cannot accurately reflect the actual noise statistical characteristics.

[0042] Based on this, in order to solve the problem of poor robustness of the time comparison tracking loop in a high dynamic scene, the application provides a signal tracking method, a modem and a time synchronization system. The method proposes a STKF tracking algorithm based on variational Bayes. The STKF algorithm not only has the advantages of robust tracking, but also adaptively adjusts the measurement noise covariance R in the STKF algorithm, improves the accuracy of the fading factor in the STKF algorithm, and further improves the robustness and accuracy of the tracking loop in a high dynamic scene. Compared with the existing method, the algorithm has principle innovation.

[0043] According to an aspect of an embodiment of the application, a signal tracking method is provided. The method can be applied to a tracking system in a modem of a time synchronization system. The tracking system uses a strong tracking algorithm filter, Figure 1 The flowchart of the signal tracking method provided by the embodiment of the application is shown in Figure 1 The method comprises the following steps:

[0044] Step S102, acquiring a time comparison signal;

[0045] In this step, the time comparison intermediate frequency signal enters the modem, and after the acquisition stage, a coarse acquisition result is obtained. Based on the coarse acquisition result, the time comparison signal of the tracking loop is obtained.

[0046] Step S104, generating a first measurement value of the carrier parameter by using the updated carrier information of the last tracking loop and the time comparison signal;

[0047] In this step, the measurement value of the carrier parameter in the current tracking loop is generated by using the updated carrier information of the last tracking loop and the time comparison signal. Referring to Figure 2 , the measurement value is the output value of the discriminator. The carrier parameter is used to update the carrier. The measurement value of the carrier parameter obtained based on the updated carrier information of the last tracking loop is taken as the first measurement value of the carrier parameter.

[0048] Step S106, calculating a measurement noise covariance matrix according to the first iteration parameter value of the strong tracking algorithm in the last tracking loop, the first estimation value of the carrier parameter in the last tracking loop and the first measurement value.

[0049] In this step, the strong tracking algorithm is iterated once in each tracking loop, and each iteration can output an estimated value of the carrier parameter. The estimated value of the carrier parameter output by the strong tracking algorithm in the last tracking loop is taken as a first estimated value, and the parameter value used by the strong tracking algorithm in the iteration in the last tracking loop is taken as a first iteration parameter value. According to the first iteration parameter value of the strong tracking algorithm in the last tracking loop, the first estimated value of the carrier parameter in the last tracking loop, and the first measurement value, a measurement noise covariance matrix is calculated by using a variational Bayesian algorithm, so that the measurement noise covariance matrix in the current tracking loop can be updated in real time, and the actual noise statistical characteristics can be more accurately reflected.

[0050] In step S108, a second estimated value of the carrier parameter in the current tracking loop is calculated by using the measurement noise covariance matrix and a second iteration parameter value of the strong tracking algorithm in the current tracking loop.

[0051] In this step, the parameter value used by the strong tracking algorithm in the iteration in the current tracking loop is taken as a second iteration parameter value, and an estimated value of the carrier parameter in the current tracking loop is calculated based on the real-time measurement noise covariance matrix and the second iteration parameter value of the strong tracking algorithm in the current tracking loop. Referring to Figure 2 , the estimated value is taken as an output of the STKF-based loop filter. The estimated value of the carrier parameter output by the strong tracking algorithm in the current tracking loop is taken as a second estimated value.

[0052] By adaptively adjusting the measurement noise covariance matrix, the estimation accuracy of the tracking loop can be improved, and the robustness and accuracy of the tracking loop in the time comparison modem in a high dynamic environment can be effectively improved.

[0053] In step S110, the carrier information is updated by using the second estimated value, and a signal tracking result of the current tracking loop is obtained.

[0054] In the embodiment of the present application, a time comparison signal is acquired; a first measurement value of a carrier parameter is generated by using updated carrier information of a previous tracking loop and the time comparison signal; a measurement noise covariance matrix is calculated according to a first iteration parameter value of a strong tracking algorithm in the previous tracking loop, a first estimated value of the carrier parameter in the previous tracking loop and the first measurement value; a second estimated value of the carrier parameter in a current tracking loop is calculated by using the measurement noise covariance matrix and a second iteration parameter value of the strong tracking algorithm in the current tracking loop; and the signal tracking result of the current tracking loop is obtained by updating the carrier information by using the second estimated value. In the embodiment of the present application, the measurement noise covariance matrix can be adaptively adjusted in each tracking loop, and the second estimated value of the carrier parameter in the strong tracking algorithm of the tracking loop can be calculated based on the adjusted measurement noise covariance matrix, so that the accuracy of the second estimated value of the carrier parameter in the strong tracking algorithm of the tracking loop can be improved, and the accuracy of the signal tracking result of the tracking loop can be improved, thereby effectively improving the robustness and accuracy of the tracking loop in the time comparison modem in a high dynamic environment.

[0055] In an alternative embodiment, the carrier parameter comprises at least a carrier phase error parameter, a carrier Doppler shift parameter and a carrier Doppler rate parameter.

[0056] In the embodiment of the present application, the carrier parameter can be a multi-dimensional vector, and the dimension of the vector can be set according to actual requirements, which is not specifically limited in the embodiment of the present application. For example, the carrier parameter can be a three-dimensional vector, comprising a carrier phase error parameter, a carrier Doppler shift parameter and a carrier Doppler rate parameter.

[0057] In an alternative embodiment, the first measurement value of the carrier parameter is generated by using the updated carrier information of the previous tracking loop and the time comparison signal, which can be executed according to the following steps:

[0058] Pseudo code information is generated; the time comparison signal is demodulated by using the updated carrier information of the previous tracking loop and the pseudo code information to obtain a demodulation result; target frequency information in the demodulation result is filtered out to obtain an integration result; the carrier phase error is calculated by using the integration result to obtain the first measurement value of the carrier parameter.

[0059] In the embodiment of the present application, after the time comparison modem receives the intermediate frequency time comparison signal, the signal is firstly subjected to coarse acquisition in the acquisition stage, and then the intermediate frequency signal enters the tracking stage as the input signal of the tracking loop. The intermediate frequency signal of the time comparison signal obtained by the tracking loop firstly enters the correlator, and is multiplied by the local carrier, i.e. the updated carrier output by the last tracking loop and the local pseudo code, so as to demodulate the time data information and obtain the demodulation result. Then, the target frequency information in the demodulation result is filtered to obtain the integration result. Finally, the integration result of the correlator is input into the carrier loop phase discriminator to obtain the first measurement value of the carrier parameter. The local pseudo code can be the pseudo code generated by the real-time pseudo code generator.

[0060] It should be noted that the output of the phase discriminator, i.e. the first measurement value of the carrier parameter, is input into the STKF tracking loop as the measurement value.

[0061] In an alternative embodiment, the measurement noise covariance matrix can be calculated according to the first iteration parameter value of the strong tracking algorithm in the last tracking loop, the first estimated value of the carrier parameter in the last tracking loop and the first measurement value, which can be performed according to the following steps:

[0062] The first degree of freedom number and the first inverse scale matrix of the measurement noise covariance matrix in the last variational Bayesian iteration are obtained; the second degree of freedom number of the measurement noise covariance matrix in the current variational Bayesian iteration is calculated by using the first degree of freedom number; the second inverse scale matrix of the measurement noise covariance matrix in the current variational Bayesian iteration is calculated by using the first inverse scale matrix, the first iteration parameter value, the first estimated value and the first measurement value; and the measurement noise covariance matrix is calculated by using the second degree of freedom number, the second inverse scale matrix and the dimension value of the first measurement value.

[0063] In the embodiment of the present application, the measurement noise covariance matrix can be updated based on the variational Bayesian algorithm, for example, the measurement noise covariance matrix in each tracking loop can be adjusted by using the variational Bayesian algorithm. In the variational Bayesian iteration, the degree of freedom and the inverse scale matrix of the measurement noise covariance matrix are updated each time the variational Bayesian iteration is updated. The degree of freedom data of the measurement noise covariance matrix in the last variational Bayesian iteration is taken as the first degree of freedom number, and the inverse scale matrix is taken as the first inverse scale matrix.

[0064] The degrees of freedom data in the current variational Bayesian iteration is taken as the second degrees of freedom number, and the inverse scale matrix is taken as the second inverse scale matrix. The second degrees of freedom number of the measurement noise covariance matrix in the current variational Bayesian iteration is calculated by using the first degrees of freedom number; the second inverse scale matrix of the measurement noise covariance matrix in the current variational Bayesian iteration is calculated by using the first inverse scale matrix, the first iteration parameter value, the first estimated value and the first measurement value. Further, the measurement noise covariance matrix is calculated by using the second degrees of freedom number, the second inverse scale matrix and the dimension value of the first measurement value.

[0065] In the standard KF algorithm, the noise covariance is always kept unchanged in the iteration process, but when the tracking loop works in a high dynamic scene, the fixed and unchanged measurement noise covariance cannot accurately reflect the noise statistical characteristics of the actual environment, therefore, the measurement noise covariance needs to be adaptively changed and adjusted in real time according to the input signal. In an optional embodiment, the measurement noise covariance matrix is calculated according to the following formula:

[0066]

[0067]

[0068]

[0069]

[0070] wherein, the degrees of freedom number is represented by d, the inverse scale matrix is represented by D, i represents the number of variational iterations, k represents the number of STKF iterations, z k the measurement vector at the k moment is represented by z, and H represents the measurement matrix, the state variable at the k moment is represented by x, k the estimated state covariance matrix is represented by P, and m represents the dimension of the measurement value vector, the measurement noise covariance matrix is represented by R.

[0071] In the embodiment of the application, the degrees of freedom number and the inverse scale matrix are updated through the variational Bayesian iteration process, and the measurement noise covariance matrix is calculated by using the updated degrees of freedom number and inverse scale matrix. In the variational Bayesian iteration process, the first iteration parameter value used can include the measurement matrix H and the estimated state covariance matrix P k The first iteration parameter value can be updated in the STKF iteration process. In the variational Bayesian iteration process of the embodiment of the application, the first measurement value used is the measurement vector z k at the k moment, and the first estimated value used is the state variable x m represents the dimension of the measurement vector, which can be determined according to the measurement vector z k is determined.

[0072] In an alternative embodiment, the second estimated value of the carrier parameter in the current tracking loop is calculated by using the measurement noise covariance matrix and the second iteration parameter value of the strong tracking algorithm in the current tracking loop, and the calculation can be performed according to the following steps:

[0073] The fading factor is calculated by using the measurement noise covariance matrix, the predicted state covariance matrix is calculated by using the fading factor and the second iteration parameter value, and the second estimated value of the carrier parameter in the current tracking loop is calculated by using the predicted state covariance matrix.

[0074] In the embodiment of the present application, the accuracy of the fading factor is calculated by adjusting the measurement noise covariance matrix in real time, considering that the accuracy of the fading factor is affected by the value of the measurement noise covariance matrix. In the specific implementation, the fading factor is calculated by using the measurement noise covariance matrix, so that the value of the fading factor is optimized, and then the predicted state covariance matrix is calculated by using the fading factor and the second iteration parameter value, and the second estimated value of the carrier parameter in the current tracking loop is calculated by using the predicted state covariance matrix.

[0075] In an alternative embodiment, the fading factor is calculated according to the following formula:

[0076] λ k = diag(λ 1,k ,λ 2,k …,λ n,k )

[0077]

[0078]

[0079]

[0080]

[0081]

[0082] wherein λ k represents the fading factor, σ i is a predefined constant, tr[·] represents the trace of a matrix, V k represents the estimated innovation covariance matrix, R k represents the measurement noise covariance matrix, H represents the measurement matrix, Q k-1 is the system noise covariance, and Φ represents the system transition matrix. Let ρ be the estimated state covariance matrix, ρ be the forgetting factor, and d be the innovation sequence.

[0083] In this embodiment of the invention, the measurement noise covariance matrix R is updated based on the variational Bayes algorithm. k This can be used as an input parameter in the STKF iteration calculation. During the STKF iteration calculation, the iterative parameter values ​​used can include the measurement matrix H and the system noise covariance Q. k-1 The system transition matrix Φ and the estimated state error covariance matrix The iteration parameters can be updated and can also be used to update other iteration parameters. The fading factor is calculated based on the iteration parameter values ​​and the measurement noise covariance matrix.

[0084] In one alternative implementation, the predicted state covariance matrix is ​​calculated according to the following formula:

[0085]

[0086] in, For the prediction state error covariance matrix, λ k Φ represents the fading factor, and Φ represents the system transition matrix. Q is the estimated state error covariance matrix. k-1 Let be the system noise covariance.

[0087] In this embodiment of the invention, after updating the fading factor λ k Next, the predicted state error covariance matrix is ​​calculated using the iterative parameter values ​​of STKF and the fading factor. The iterative parameter values ​​used may include: the system transition matrix Φ, and the estimated state error covariance matrix. and system noise covariance Q k-1 .

[0088] The implementation of this method will be described below with a specific example.

[0089] See Figure 2 The diagram shown illustrates a STKF tracking loop based on variational Bayesian methods. The intermediate frequency (IF) time comparison signal undergoes an acquisition stage to obtain a coarse acquisition result, and then enters the tracking stage. The signal entering the tracking stage is divided into two paths: one is multiplied by a sinusoidal local carrier (path I), and the other is multiplied by a cosine local carrier (path Q). Subsequently, both signals multiplied by the local carrier are multiplied by local pseudocode to extract timing information. After multiplication by the pseudocode, both signals are input to an integrator clearer to filter out high-frequency components and retain low-frequency components, thus obtaining two integral values ​​I. p (k) and Q p(k), the integral value is input into the phase discriminator value, and carrier phase error and other carrier parameter information are obtained. The obtained carrier phase error and other carrier parameter information are input into the variational Bayesian algorithm and the STKF tracking loop, and the variational Bayesian algorithm is used to adjust the STKF carrier tracking loop.

[0090] The variational Bayesian method is introduced to adaptively adjust the measurement noise covariance R, and the updated R is used to calculate the fading factor, so that the accuracy of the fading factor can be improved. In the STKF algorithm, a fading factor needs to be introduced to adjust the state error covariance matrix. The calculated fading factor is applied to the calculation of the state error covariance matrix, so as to adjust the Kalman gain, and then enter the iterative filtering of the next moment until the accurate carrier phase error and other carrier parameters are obtained. The estimated carrier phase and carrier Doppler frequency are used to update the Doppler frequency and carrier phase in the carrier numerically controlled oscillator (NCO), so as to obtain updated carrier information and signal tracking results of the tracking loop. The present application introduces the variational Bayesian algorithm to adaptively adjust R, so that R can be adaptively adjusted according to the change of the external environment, thereby improving the accuracy of the fading factor in the STKF algorithm, and further improving the estimation accuracy of the tracking loop, and effectively improving the robustness and accuracy of the tracking loop in the time comparison modem in the high dynamic environment.

[0091] Figure 3 The algorithm flowchart of the variational Bayesian STKF tracking loop is shown, and the tracking loop based on the variational Bayesian STKF algorithm is used. Specifically, the following steps can be performed: the captured intermediate frequency time comparison signal is input into the tracking loop, the intermediate frequency signal enters the correlator and is multiplied by the local carrier and the local pseudo code to obtain the integral value, the integral value is input into the phase discriminator, the carrier phase error is obtained, the carrier phase error is input into the STKF tracking loop as a measurement value, the R is updated by the variational Bayesian method, and the and α k , the Doppler frequency and the carrier phase are updated according to ω k and α k .

[0092] The principles of the filtering process of the STKF adjusted by the variational Bayesian method will be described below according to steps one to step five.

[0093] Step one: system modeling

[0094] In the discrete Kalman filter, the state model and the measurement model are respectively:

[0095] x k =Φ k x k-1 +n k-1

[0096] z k = H k x k + v k

[0097] where x k represents the state variable at time k, Φ k represents the system transition matrix, z k+1 represents the measurement vector at time k+1, H k represents the measurement matrix, n k and v k represent the system noise matrix and the measurement noise matrix, respectively, n k and v k are mutually independent white noise sequences;

[0098] For a third-order tracking loop based on KF, the state model and the measurement model are as follows:

[0099]

[0100]

[0101] where, represents the carrier phase error, with the unit of radian (rad), ω k represents the carrier Doppler shift, with the unit of rad / s, α k represents the carrier Doppler rate, with the unit of rad / s 2 . T represents the update period of the carrier tracking loop, with the unit of second (s). is the measurement value, which is usually the carrier phase error value output by the phase discriminator. n k represents the system noise, v k represents the measurement noise.

[0102] Step 2: Standard KF (Kalman Filter) iteration process

[0103] In the iteration process, first, the state vector x and the state prediction covariance matrix P are initialized. At the same time, the system noise covariance Q (calculated based on the system noise in step 1) and the measurement noise covariance R (calculated based on the measurement noise in step 1) are initialized.

[0104] Then, the KF filtering iteration process is performed, which includes two parts, the prediction process and the update process, and the specific formula is as follows:

[0105]

[0106]

[0107]

[0108]

[0109]

[0110] where, is the estimated state value, is the estimated state error covariance matrix, is the predicted state estimate, is the predicted state error covariance matrix, K k is the Kalman gain.

[0111] Step three: update the measurement noise covariance R using the variational Bayes method

[0112] In the KF algorithm, the state vector x k , the measurement covariance matrix R k are the values to be estimated, and the two quantities are independent of each other. The purpose of KF filtering is to find the posterior distribution p(x k , R k |z 1:k ). According to the variational Bayes algorithm, under the condition of known observation at the first k time and initial value p(x0, R0), the posterior distribution p(x k , R k |z 1:k ) is estimated through the prediction and update processes.

[0113] Assume that the joint probability distribution of x k-1 , R k-1 at time k-1 is:

[0114] p(x k-1 , R k-1 |z 1:k-1 )

[0115] =N(x k-1 |m k-1 , P k ) × IW(R k-1 |v k-1 , V k-1 )

[0116] where m k-1 represents the mean of x k-1 , IW() represents the inverse Wishart distribution, v represents the number of degrees of freedom, V represents the inverse scale matrix, and N() represents the Gaussian distribution.

[0117] The prediction process is:

[0118] p(xk ,R k |z 1:k-1 )

[0119] =p(x k |z 1:k-1 )p(R k |z 1:k-1 )

[0120] =N(x) k-1 |x k|k-1 ,P k|k-1 )IW(R k |v k|k-1 V k|k-1 )

[0121] Where, x k|k-1 P k|k-1 These are the one-step predicted values ​​of the state vector and the state error covariance matrix, respectively. k|k-1 V k|k-1 This is a one-step prediction of the parameters of the inverse Wishart distribution.

[0122] The update process is as follows:

[0123] Posterior distribution p(x) k ,R k |z 1:k It can be approximated as:

[0124] p(x k ,R k |z 1:k )≈Q(x k )Q(R k )

[0125] This approximate distribution can be obtained by calculating the Kullback-Leibler (KL) divergence between the true distribution and the approximate distribution, i.e.

[0126]

[0127] Where Q(x) k ) and Q(R k That is, p(x) is the value that needs to be solved. k ,R k |z 1:k The approximate distribution of Q(x). k Q(R) follows a Gaussian distribution. k It follows an inverse Wishart distribution, that is:

[0128] Q(x k )=N(x k |x k ,P k )

[0129] Q(R k )=IW(R k |v k ,V k )

[0130] The measurement noise covariance R can be updated in real time through the above prediction and update process.

[0131] Therefore, according to the above derivation, the process of adjusting R is summarized as follows:

[0132]

[0133]

[0134]

[0135]

[0136] Then, the adjusted R is applied to the . Where i represents the number of times of variational iteration, and m represents the dimension of the measurement vector.

[0137] Step four: calculate the fading factor

[0138] In the STKF algorithm, a fading factor needs to be introduced to adjust the state prediction covariance matrix, that is,

[0139]

[0140] The fading factor λ k = diag(λ 1,k ,λ 2,k …,λ n,k ) is given by:

[0141]

[0142]

[0143] Where N k and M k are defined as:

[0144]

[0145]

[0146]

[0147] Where V k represents the estimated information covariance matrix. tr[] represents the trace of the matrix. ρ is the forgetting factor, and σi is a predefined constant, d represents an information sequence.

[0148] Therefore, the accuracy of the fading factor is affected by the value of R. The STKF algorithm based on variational Bayes proposed in the application adjusts the measurement noise covariance matrix R in real time, thereby improving the accuracy of the fading factor.

[0149] Step five: applying the fading factor

[0150] The calculated fading factor is applied to , thereby adjusting the Kalman gain.

[0151] Then, the iteration filtering of the next moment is entered until the accurate carrier phase error and the like are obtained.

[0152] The estimated carrier phase error and the like are used to update the carrier NCO, thereby performing the next tracking loop update.

[0153] The application provides a signal tracking method, a modem and a time synchronization system, which improves the robustness and accuracy of the tracking loop in the time comparison modem in a high dynamic scene. Since the modem is the core equipment of the time synchronization system, the application can improve the performance of the time synchronization system in a high dynamic scene. The method innovates in principle and has strong technical disruptive power. It is expected to greatly improve the stability of the time synchronization system in a high dynamic platform and has great and far-reaching practical application significance for improving the time synchronization accuracy of the high platform in the field of aerospace and navigation.

[0154] According to another aspect of the embodiments of the application, a signal tracking device is also provided, Figure 4 As shown in the schematic diagram of the signal tracking device provided by the embodiments of the application, Figure 4 the signal tracking device comprises an acquisition module 40, a generation module 42, a calculation module 44, an estimation module 46 and an update module 48. The signal tracking device will be described in detail below.

[0155] The acquisition module 40 is configured to acquire a time comparison signal; the generation module 42 is configured to generate a first measurement value of a carrier parameter by using updated carrier information of a last tracking loop and the time comparison signal; the calculation module 44 is configured to calculate a measurement noise covariance matrix according to a first iteration parameter value of a strong tracking algorithm in the last tracking loop, a first estimation value of the carrier parameter in the last tracking loop and the first measurement value; the estimation module 46 is configured to calculate a second estimation value of the carrier parameter in a current tracking loop by using the measurement noise covariance matrix and a second iteration parameter value of the strong tracking algorithm in the current tracking loop; and the update module 48 is configured to update the carrier information by using the second estimation value, so as to obtain a signal tracking result of the current tracking loop.

[0156] In the embodiment of the present application, the time comparison signal is acquired, the first measurement value of the carrier parameter is generated by using the updated carrier information of the last tracking loop and the time comparison signal, the measurement noise covariance matrix is calculated according to the first iteration parameter value of the strong tracking algorithm in the last tracking loop, the first estimated value of the carrier parameter in the last tracking loop and the first measurement value, the second estimated value of the carrier parameter in the current tracking loop is calculated by using the measurement noise covariance matrix and the second iteration parameter value of the strong tracking algorithm in the current tracking loop, and the signal tracking result of the current tracking loop is obtained by updating the carrier information by using the second estimated value. In each tracking loop, the measurement noise covariance matrix can be adaptively adjusted in the embodiment of the present application, and the second estimated value of the carrier parameter in the strong tracking algorithm of the tracking loop can be improved based on the adjusted measurement noise covariance matrix, so as to improve the accuracy of the signal tracking result of the tracking loop, and further effectively improve the robustness and accuracy of the tracking loop in the time comparison modem in a high dynamic environment.

[0157] It should be noted that the acquisition module 40, the generation module 42, the calculation module 44, the estimation module 46 and the update module 48 correspond to steps S102 to S110 in the method embodiment, and the above modules have the same examples and application scenarios as the corresponding steps, but are not limited to the content disclosed in the above method embodiment.

[0158] In an optional embodiment, the carrier parameter at least includes a carrier phase error parameter, a carrier Doppler shift parameter and a carrier Doppler rate parameter.

[0159] In an optional embodiment, the first measurement value of the carrier parameter is generated by using the updated carrier information of the last tracking loop and the time comparison signal, including: generating pseudo code information; demodulating the time comparison signal by using the updated carrier information of the last tracking loop and the pseudo code information to obtain a demodulation result; filtering the target frequency information in the demodulation result to obtain an integration result; calculating the carrier phase error by using the integration result to obtain the first measurement value of the carrier parameter.

[0160] In an alternative implementation, the measurement noise covariance matrix is calculated according to the first iteration parameter value of the strong tracking algorithm in the last tracking loop, the first estimated value of the carrier parameter in the last tracking loop and the first measurement value, comprising: obtaining the first number of degrees of freedom and the first inverse scale matrix of the measurement noise covariance matrix in the last variational Bayesian iteration; calculating the second number of degrees of freedom of the measurement noise covariance matrix in the current variational Bayesian iteration by using the first number of degrees of freedom; calculating the second inverse scale matrix of the measurement noise covariance matrix in the current variational Bayesian iteration by using the first inverse scale matrix, the first iteration parameter value, the first estimated value and the first measurement value; calculating the measurement noise covariance matrix by using the second number of degrees of freedom, the second inverse scale matrix and the dimension value of the first measurement value.

[0161] In an alternative implementation, the measurement noise covariance matrix is calculated according to the following formula:

[0162]

[0163]

[0164]

[0165]

[0166] wherein, denotes the number of degrees of freedom, denotes the inverse scale matrix, i denotes the number of variational iterations, k denotes the number of STKF iterations, z k denotes the measurement vector at time k, H represents the measurement matrix, denotes the state variable at time k, P k is the estimated state error covariance matrix, m denotes the dimension of the measurement vector, denotes the measurement noise covariance matrix.

[0167] In an alternative implementation, the second estimated value of the carrier parameter in the current tracking loop is calculated by using the measurement noise covariance matrix and the second iteration parameter value of the strong tracking algorithm in the current tracking loop, comprising: calculating the fading factor by using the measurement noise covariance matrix; calculating the predicted state covariance matrix by using the fading factor and the second iteration parameter value; calculating the second estimated value of the carrier parameter in the current tracking loop by using the predicted state error covariance matrix.

[0168] In an alternative implementation, the fading factor is calculated according to the following formula:

[0169] λ k = diag(λ 1,k ,λ2,k …,λ n,k )

[0170]

[0171]

[0172]

[0173]

[0174]

[0175] where λ k represents a fading factor, σ i is a predefined constant, tr[·] represents taking the trace of a matrix, V k represents an estimated innovation covariance matrix, R k represents a measurement noise covariance matrix, H represents a measurement matrix, Q k-1 is a system noise covariance, Φ represents a system transition matrix, is an estimated state error covariance matrix, ρ is a forgetting factor, and d represents an innovation sequence.

[0176] In an alternative embodiment, the predicted state error covariance matrix is calculated according to the following equation:

[0177]

[0178] where P is the predicted state error covariance matrix, λ k represents a fading factor, Φ represents a system transition matrix, is an estimated state error covariance matrix, Q k-1 is a system noise covariance.

[0179] According to another aspect of an embodiment of the present application, there is also provided a modem comprising: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to perform the above-mentioned signal tracking method.

[0180] According to another aspect of an embodiment of the present application, there is also provided a time synchronization system comprising the above-mentioned modem.

[0181] The above merely provides a preferred embodiment of the present application, and is not intended to limit the protection scope of the present application.

Claims

1. A signal tracking method, characterized by, The method comprises: acquiring a time comparison signal; generating a first measurement value of a carrier parameter by using updated carrier information of a previous tracking loop and the time comparison signal; calculating a measurement noise covariance matrix according to a first iteration parameter value of a strong tracking algorithm in the previous tracking loop, a first estimated value of the carrier parameter in the previous tracking loop, and the first measurement value; calculating a second estimated value of the carrier parameter in a current tracking loop by using the measurement noise covariance matrix and a second iteration parameter value of the strong tracking algorithm in the current tracking loop; updating the carrier information by using the second estimated value to obtain a signal tracking result of the current tracking loop; calculating a measurement noise covariance matrix according to a first iteration parameter value of a strong tracking algorithm in a previous tracking loop, a first estimated value of a carrier parameter in the previous tracking loop, and a first measurement value, comprising: acquiring a first degree of freedom number and a first inverse scale matrix of the measurement noise covariance matrix in a previous variational Bayes iteration; calculating a second degree of freedom number of the measurement noise covariance matrix in a current variational Bayes iteration by using the first degree of freedom number; calculating a second inverse scale matrix of the measurement noise covariance matrix in the current variational Bayes iteration by using the first inverse scale matrix, the first iteration parameter value, the first estimated value, and the first measurement value; calculating the measurement noise covariance matrix by using the second degree of freedom number, the second inverse scale matrix, and a dimension value of the first measurement value; calculating the measurement noise covariance matrix according to the following formula: wherein, denotes the number of degrees of freedom, denotes the inverse scale matrix, i denotes the number of variational iterations, k denotes the number of STKF iterations, z k denotes the measurement vector at time k, H represents the measurement matrix, denotes the state variable at time k, P k is the estimated state covariance matrix, m denotes the dimension of the measurement vector, denotes the measurement noise covariance matrix.

2. The method of claim 1, wherein, The carrier parameter at least comprises a carrier phase error parameter, a carrier Doppler shift parameter, and a carrier Doppler rate parameter.

3. The method of claim 1, wherein, The method for generating a first measurement value of a carrier parameter by using updated carrier information of a previous tracking loop and a time comparison signal comprises: generating pseudo-code information; demodulating the time comparison signal by using the updated carrier information of the previous tracking loop and the pseudo-code information to obtain a demodulation result; filtering out target frequency information in the demodulation result to obtain an integration result; calculating a carrier phase error by using the integration result to obtain the first measurement value of the carrier parameter.

4. The method of claim 1, wherein, The method for calculating a second estimated value of a carrier parameter in a current tracking loop by using a measurement noise covariance matrix and a second iteration parameter value of a strong tracking algorithm in the current tracking loop comprises: calculating a fading factor by using the measurement noise covariance matrix; calculating a predicted state covariance matrix by using the fading factor and the second iteration parameter value; calculating the second estimated value of the carrier parameter in the current tracking loop by using the predicted state covariance matrix.

5. The method of claim 4, wherein, The method comprises: calculating the fading factor according to the following formula: λ k = diag(λ 1,k , λ 2,k ,..., λ n,k ) where λ k represents the fading factor, σ i is a predefined constant, tr[·] denotes the trace of a matrix, V k represents the estimated innovation covariance matrix, R k stands for the measurement noise covariance matrix, H represents the measurement matrix, Q k-1 is the system noise covariance, Φ represents the system transition matrix, is the estimated state covariance matrix, p is the forgetting factor, d represents the innovation sequence.

6. The method of claim 4, wherein, The method comprises: calculating the predicted state covariance matrix according to the following formula: wherein, is the predicted state covariance matrix, λ k denotes a fading factor, Φ represents the system transition matrix, is the estimated state covariance matrix, Q k-1 is the system noise covariance.

7. A modem, characterized by The method comprises: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to execute the signal tracking method in any one of claims 1 to 6.

8. A time synchronization system, characterized by The modem comprises the method in claim 7. The modem comprises the method in claim 7.

Images