CN114236577A - GNSS signal capturing method based on artificial neural network - Google Patents

GNSS signal capturing method based on artificial neural network Download PDF

Info

Publication number
CN114236577A
CN114236577A CN202111553951.8A CN202111553951A CN114236577A CN 114236577 A CN114236577 A CN 114236577A CN 202111553951 A CN202111553951 A CN 202111553951A CN 114236577 A CN114236577 A CN 114236577A
Authority
CN
China
Prior art keywords
neural network
artificial neural
correlation peak
network model
frequency
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202111553951.8A
Other languages
Chinese (zh)
Other versions
CN114236577B (en
Inventor
王永
罗兵
孙娟娟
邱冬悦
李祥杰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shandong University
Original Assignee
Shandong University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shandong University filed Critical Shandong University
Priority to CN202111553951.8A priority Critical patent/CN114236577B/en
Publication of CN114236577A publication Critical patent/CN114236577A/en
Application granted granted Critical
Publication of CN114236577B publication Critical patent/CN114236577B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
    • G01S19/01—Satellite radio beacon positioning systems transmitting time-stamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
    • G01S19/13—Receivers
    • G01S19/24—Acquisition or tracking or demodulation of signals transmitted by the system
    • G01S19/29—Acquisition or tracking or demodulation of signals transmitted by the system carrier including Doppler, related
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
    • G01S19/01—Satellite radio beacon positioning systems transmitting time-stamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
    • G01S19/13—Receivers
    • G01S19/24—Acquisition or tracking or demodulation of signals transmitted by the system
    • G01S19/30—Acquisition or tracking or demodulation of signals transmitted by the system code related
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • General Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computing Systems (AREA)
  • Molecular Biology (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Feedback Control In General (AREA)
  • Position Fixing By Use Of Radio Waves (AREA)

Abstract

The invention relates to a GNSS signal capturing method based on an artificial neural network, which trains a multilayer perceptron neural network by taking maximum correlation peak value generated during GNSS signal capturing and maximum correlation peak value data of nearby frequency points as a data set to obtain an optimal neural network structure and parameter values, and carries out high-precision carrier frequency prediction according to the obtained neural network structure and parameters. The invention predicts the spatial distribution of the related peak values by using the artificial neural network, shortens the search time of carrier frequency, realizes the rapid capture of signals, and improves the capture precision and capture speed.

Description

GNSS signal capturing method based on artificial neural network
Technical Field
The invention relates to a GNSS signal capturing method based on an artificial neural network, and belongs to the technical field of satellite positioning and navigation.
Background
The satellite navigation positioning system is widely applied to the aspects of vehicle navigation, aviation and navigation, geographic mapping, mass consumption and the like, can provide position and time information for users, and realizes positioning and navigation functions for terminal users by transmitting radio navigation signals through a plurality of navigation satellites in space.
The receiver is the core part of the navigation positioning system and generally consists of an antenna, a radio frequency front end and a baseband signal processing part. Receiving satellite signals broadcast by visible satellites in a space constellation by an antenna; the radio frequency front end converts radio frequency signals received by an antenna into digital intermediate frequency signals which are easy to process through amplification, down conversion, filtering and A/D sampling quantification, and then the digital intermediate frequency signals are sent to a baseband signal processing part; the baseband signal processing part acquires pseudo-range information by capturing and tracking the digital intermediate frequency signal, demodulates satellite position information, satellite running state information, clock correction information, ionosphere correction information and the like contained in the signal, and finally calculates the position of the receiver. The acquisition is a core step of baseband signal processing, and is used as an initial part of signal synchronization of the navigation positioning receiver, and the performance of the acquisition directly influences the precision and the processing speed of a subsequent tracking loop. The acquisition process includes demodulation and correlation processing of the signal, and the main tasks are to identify the current receiver visible satellites, to obtain roughly the received signal carrier frequency and code phase, and to provide parameter estimation for the subsequent tracking module. Therefore, how to improve the acquisition performance becomes a hot issue in the field of satellite positioning and navigation. The acquisition of satellite signals is in fact a three-dimensional search process with respect to the visible satellites, the carrier frequency and the pseudo-random code phase, the receiver having to perform the maximum two-dimensional search of the respective satellite signals at start-up. The maximum doppler frequency shift amount caused by the relative motion of the user receiver and the satellite in the connection direction is ± 10KHz, and the 20KHz indefinite interval with the carrier nominal frequency f as the center is usually used as the frequency search range for capturing the satellite signal when the receiver is started. After determining the search range of the doppler shift of the signal, the receiver needs to search sequentially from the initial value of the search range of the frequency with a certain search step length until the signal is detected finally or all frequency ranges are searched. When the step setting of carrier frequency searching is smaller, the error of frequency is smaller, but the related calculation amount is increased, so that the acquisition time is longer, and the performance of a receiver is influenced; when the step length of carrier frequency search is set to be larger, the error of frequency is larger, the signal component output by the correlator is weaker, the alarm missing rate of signal detection is increased, the sensitivity of signal capture is reduced, and when the estimation precision of the carrier frequency is overlarge, the dynamic adjustment burden of a tracking loop is increased, the demodulation time of navigation data is increased, the performance of a receiver is influenced, and in severe cases, the tracking loop cannot pull the signal to a locking state, so that the navigation data is demodulated mistakenly.
In the implementation of the conventional acquisition algorithm, in order to meet the acquisition accuracy problem, the frequency domain search is mostly completed, then the search step length is reduced in a small range, and the search is repeated for a plurality of times, the estimation accuracy of the carrier frequency is dozens of hertz, the acquisition time is increased by the method, and the high-accuracy carrier frequency is difficult to obtain. Therefore, it is urgently needed to design a capturing method with good performance, so as to reduce the time for capturing and searching while acquiring a high-precision carrier frequency.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a GNSS signal capturing method based on an artificial neural network.
The core idea of the invention is to combine the parallel code phase search capture algorithm with the artificial neural network optimization algorithm, and through training the artificial neural network model, the accurate carrier frequency is predicted quickly, thereby greatly reducing the capture time, considering the capture time and the capture accuracy and optimizing the performance of the receiver.
Interpretation of terms:
1. GNSS signals, GNSS is an abbreviation of Global Navigation Satellite System, called Global Navigation Satellite System signals or Global Navigation Satellite System signals.
2. PRN, is an abbreviation of pseudo random noise code (pseudo random noise code).
3. The Levenberg-Marquardt algorithm, is a Levenberg-Marquardt algorithm, and is called LM algorithm for short. The least square fitting for calculating the nonlinear function has wide application in parameter fitting, is an algorithm for solving the maximum (small) value by using a gradient, and has the advantages of a gradient method and a Newton method.
The technical scheme of the invention is as follows:
a GNSS signal capturing method based on an artificial neural network comprises the following concrete implementation steps:
step A: acquiring a data set comprising:
step 1: obtaining a correlation peak value in a time domain;
step 2: detecting the correlation peak value in the time domain obtained in the step 1 by adopting an average correlation peak detection method, and judging whether the current satellite is captured or not; the method comprises the following steps: if the ratio of the maximum correlation peak value to the average correlation peak value is larger than a set threshold value, entering the step 3; otherwise, entering step 4; wherein, the maximum correlation peak value refers to the maximum value in the correlation peak values in the time domain obtained in step 1, and the average correlation peak value is the average value of the correlation peak values in the time domain obtained in step 1;
and step 3: reducing the frequency searching step length, executing step 1, judging whether the carrier frequency searching is finished, and if the carrier frequency searching is finished, intercepting the maximum correlation peak value of the current frequency point and the maximum correlation peak value of the frequency points nearby as a data set; the data set comprises a training set and a test set; otherwise, continuing to execute the step 1;
and 4, step 4: judging whether the carrier frequency searching is finished or not, if so, determining that the current satellite PRN is not captured, adjusting the next satellite PRN, and performing the step 5, otherwise, performing the step 1;
and 5: judging whether the satellite PRN search is finished, if so, finishing the acquisition of the GNSS signal, otherwise, executing the step 1;
and B: training an artificial neural network model, comprising:
step 6: inputting the training set obtained in the step A into an artificial neural network model for training and maximum value prediction;
the method specifically comprises the following steps: training an artificial neural network model, wherein the training cutoff condition is that iteration times are finished or errors meet requirements, and obtaining an optimal neural network structure and parameters, namely the trained artificial neural network model, according to a local optimal principle;
inputting test data in the test set into the trained artificial neural network model, testing the detection effect of the trained artificial neural network model, and predicting the position of the maximum value;
and C: GNSS signal capture is carried out through a trained artificial neural network model;
during testing, the related peak test set data generated during GNSS signal capturing is obtained through the steps 1-5, the test set data is input into a trained artificial neural network model, calculation is carried out in the artificial neural network model along the data flowing direction until the data is transmitted to an output layer and output, one-time prediction is completed, and the GNSS signal capturing is realized.
The training of the multilayer perceptron neural network is based on a Levenberg-Marquardt learning algorithm to obtain the optimal neural network structure and parameter values, and the capturing accuracy and precision are improved.
According to the invention, the step 1 is preferably realized by the following steps:
step 1.1: mixing the received digital intermediate frequency signal with a copy sine carrier signal and a copy cosine carrier signal of a certain frequency respectively to obtain a baseband complex signal;
step 1.2: carrying out Fourier transform on the baseband complex signal obtained in the step 1.1;
step 1.3: carrying out Fourier transform on the local copy pseudo code, and multiplying the conjugate value after Fourier transform by the result obtained in the step 1.2;
step 1.4: performing inverse Fourier transform on the result obtained in the step 1.3;
step 1.5: and (4) carrying out modulus square on the result obtained in the step 1.4 to obtain a correlation peak value in a time domain.
Preferably, in step 2, an average correlation peak detection method is used to detect the correlation peak value in the time domain obtained in step 1, specifically:
maximum correlation peak value apeak＝max(A)；
Mean correlation peak
Figure BDA0003418007300000031
N represents the length of the sequence subjected to fourier transform; a is a two-dimensional array which is generated after modular squaring and takes a pseudo-random code phase index value as an X axis and takes a correlation peak value corresponding to each pseudo-random code phase index value as a Y axis;
Figure BDA0003418007300000032
where m denotes the size of the frequency index value, fMPRepresenting the maximum Doppler shift, fIFIndicating the nominal carrier frequency, f, of the digital intermediate frequency signalbinRepresenting a carrier frequency search step;
n represents the magnitude of the pseudo-code phase index value, t represents the pseudo-code length, tcRepresenting the maximum pseudo-code phase shift, tbinSearching a step size for the pseudo code phase;
o denotes the size of the frequency index value after reducing the frequency search step size, fCRepresenting the center carrier frequency, f, for artificial neural network model predictionMCMaximum search Range, f, representing Artificial neural network model predictioncminRepresenting the carrier frequency search step predicted by the artificial neural network model, and N representing the satellite PRN sequence to be acquired.
According to the present invention, in step 2, the maximum correlation peak of the current frequency point and the maximum correlation peak of the frequency points in the vicinity thereof are intercepted as the data sets, specifically: the current frequency point is taken as the center, the left side and the right side are respectively extended by p frequency points, and the maximum correlation peak value A corresponding to each frequency point is obtainedpeakComposed search matrix AsAs a set of data, it is possible to,
Figure BDA0003418007300000041
floor is a floor rounding function.
According to the optimization method, the artificial neural network model comprises an input layer, a hidden layer and an output layer, the optimization algorithm uses a Levenberg-Marquardt optimization algorithm, the hidden layer uses a Sigmoid activation function, and the output layer uses a linear activation function.
Preferably, in step 6, the training process of the artificial neural network model is as follows:
adopting a supervised learning training mode, after a training set is input into an input layer, enabling each neuron to flow into a corresponding neuron in the next layer, summing and transmitting in a hidden layer, and finally, enabling the training set to reach an output layer for processing;
once the artificial neural network model calculates the output corresponding to one of the inputs, a loss function calculates an error vector, the loss function being a mean square error function: as shown in formula (I):
Figure BDA0003418007300000042
in formula (I), x is the input vector in the training set, y (x) is the output generated by the artificial neural network, y is the desired output, n is the size of the training set, w is the weight vector, b is the bias;
obtaining gradients of all parameters of the artificial neural network model by using a back propagation algorithm, and updating all parameters of the artificial neural network model by using a Levenberg-Marquardt optimization algorithm;
and finishing the training when the loss function converges to a certain degree or the iteration times are finished, and storing the parameters of the trained artificial neural network model to obtain the trained artificial neural network model.
A computer device comprising a memory storing a computer program and a processor implementing the steps of an artificial neural network based GNSS signal acquisition method when executing the computer program.
A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the artificial neural network-based GNSS signal acquisition method.
The invention has the beneficial effects that:
1. the invention relates to a GNSS signal capturing method based on an artificial neural network, which utilizes the artificial neural network to express the nonlinear relation between the maximum correlation peak value and the carrier frequency, predicts the spatial distribution of the correlation peak value according to the relation, shortens the search time of the carrier frequency, realizes the rapid capturing of signals and improves the processing efficiency of a receiver.
2. The invention relates to a GNSS signal capturing method based on an artificial neural network, which trains a multilayer perceptron neural network based on an artificial neural network function, obtains an optimal network structure and parameter values according to a local optimal principle, and improves capturing accuracy and precision.
3. The GNSS signal capturing method based on the artificial neural network combines a parallel code phase search capturing algorithm and the artificial neural network technology, simultaneously gives consideration to capturing precision and capturing efficiency, and realizes fast and high-precision GNSS signal capturing.
Drawings
FIG. 1 is a flow chart illustrating the present invention for determining correlation peak in the time domain;
FIG. 2 is a schematic overall flowchart of an artificial neural network-based GNSS signal capturing method according to the present invention;
FIG. 3 is a schematic structural diagram of an artificial neural network model according to the present invention.
Detailed Description
The invention is further described, but not limited to, in the following description, in conjunction with the drawings and examples:
example 1
As shown in figure 2, correlation peak values of search units with different frequencies on the same code band are sampling points of a function curve of a sine function, the multi-layer perceptron neural network is trained by taking the maximum correlation peak value generated during GNSS signal capturing and the maximum correlation peak value data of nearby frequency points as data sets based on the artificial neural network function, and the optimal neural network structure and parameter values are obtained. And carrying out high-precision carrier frequency prediction aiming at the obtained neural network structure and parameters. The concrete implementation steps comprise:
step A: acquiring a data set comprising:
step 1: obtaining a correlation peak value in a time domain;
the correlation peak in the time domain refers to: the two sequences x (n) and Y (n) are correlated in the time domain, corresponding to their Fourier transforms X (k) and Y*(k)(Y*(k) Is the conjugate of Y (k) is multiplied in the frequency domain by X (k) Y*(k) The inverse fourier transform of (a) is just the correlation value at each code phase that needs to be detected. And performing modular squaring on the result of the Fourier inverse transformation, and finding out the maximum value of the sequence after the modular squaring so as to obtain the correlation peak value in the time sequence.
Step 2: detecting the correlation peak value in the time domain obtained in the step 1 by adopting an average correlation peak detection method, and judging whether the current satellite is captured or not; the method comprises the following steps: if the ratio of the maximum correlation peak value to the average correlation peak value is larger than a set threshold value, the set threshold value is related to the distribution probability of the correlation value, and when the satellite signal is stronger, the value of the threshold value is 25; when the satellite signal is weak, the threshold value is 15. Entering the step 3; otherwise, entering step 4; wherein, the maximum correlation peak value refers to the maximum value in the correlation peak values in the time domain obtained in step 1, and the average correlation peak value is the average value of the correlation peak values in the time domain obtained in step 1;
and step 3: reducing the frequency searching step length, executing step 1, judging whether the carrier frequency searching is finished, and if the carrier frequency searching is finished, intercepting the maximum correlation peak value of the current frequency point and the maximum correlation peak value of the frequency points nearby as a data set; the data set comprises a training set and a test set; otherwise, continuing to execute the step 1;
and 4, step 4: judging whether the carrier frequency searching is finished or not, if so, determining that the current satellite PRN is not captured, adjusting the next satellite PRN, and performing the step 5, otherwise, performing the step 1;
and 5: judging whether the satellite PRN search is finished, if so, finishing the acquisition of the GNSS signal, otherwise, executing the step 1;
and B: training an artificial neural network model, comprising:
step 6: inputting the training set obtained in the step A into an artificial neural network model for training and maximum value prediction; the structure of the artificial neural network model is preset according to the length of the intercepted data set;
the method specifically comprises the following steps: training an artificial neural network model, wherein the training cutoff condition is that iteration times are finished or errors meet requirements, and obtaining an optimal neural network structure and parameters, namely the trained artificial neural network model, according to a local optimal principle;
inputting test data in the test set into the trained artificial neural network model, testing the detection effect of the trained artificial neural network model, and predicting the position of the maximum value;
and C: GNSS signal capture is carried out through a trained artificial neural network model;
during testing, the related peak test set data generated during GNSS signal capturing is obtained through the steps 1-5, the test set data is input into a trained artificial neural network model, calculation is carried out in the artificial neural network model along the data flowing direction until the data is transmitted to an output layer and output, one-time prediction is completed, and the GNSS signal capturing is realized.
Example 2
The GNSS signal capturing method based on the artificial neural network according to embodiment 1 is characterized in that:
as shown in fig. 1, the specific implementation steps of step 1 include:
step 1.1: mixing the received digital intermediate frequency signal with a copy sine carrier signal and a copy cosine carrier signal of a certain frequency respectively to obtain a baseband complex signal;
step 1.2: carrying out Fourier transform on the baseband complex signal obtained in the step 1.1;
step 1.3: carrying out Fourier transform on the local copy pseudo code, and multiplying the conjugate value after Fourier transform by the result obtained in the step 1.2;
step 1.4: performing inverse Fourier transform on the result obtained in the step 1.3;
step 1.5: and (4) carrying out modulus square on the result obtained in the step 1.4 to obtain a correlation peak value in a time domain.
In step 2, an average correlation peak detection method is adopted to detect the correlation peak value in the time domain obtained in step 1, specifically:
maximum correlation peak value apeak＝max(A)；
Mean correlation peak
Figure BDA0003418007300000061
N represents the length of the sequence subjected to fourier transform; a is a two-dimensional array which is generated after modular squaring and takes a pseudo-random code phase index value as an X axis and takes a correlation peak value corresponding to each pseudo-random code index value as a Y axis; x has a size of 2NThe correlation peak values are distributed in space as a function curve of | sinc |, and the artificial neural network is used for predicting the frequency index value corresponding to the maximum correlation peak value, so that the carrier frequency required to be captured can be obtained.
Figure BDA0003418007300000071
Where m denotes the size of the frequency index value, fMPRepresenting the maximum Doppler shift, fIFIndicating the nominal carrier frequency, f, of the digital intermediate frequency signalbinRepresenting a carrier frequency search step;
n represents the magnitude of the pseudo-code phase index value, t represents the pseudo-code length, tcRepresenting the maximum pseudo-code phase shift, tbinSearching a step size for the pseudo code phase;
o denotes the size of the frequency index value after reducing the frequency search step size, fCRepresenting the center carrier frequency, f, for artificial neural network model predictionMCMaximum search Range, f, representing Artificial neural network model predictioncminRepresenting the carrier frequency search step predicted by the artificial neural network model, and N representing the satellite PRN sequence to be acquired.
In step 2, intercepting the maximum correlation peak value of the current frequency point and the maximum correlation peak value of the frequency points nearby the current frequency point as a data set, specifically: centering on the current frequency pointExtending p frequency points left and right to obtain maximum correlation peak value A corresponding to each frequency pointpeakComposed search matrix AsAs a set of data, it is possible to,
Figure BDA0003418007300000072
floor is a floor rounding function.
As shown in fig. 3, the artificial neural network model includes an input layer, a hidden layer and an output layer, the optimization algorithm uses a Levenberg-Marquardt optimization algorithm, the hidden layer uses a Sigmoid activation function, and the output layer uses a linear activation function.
In step 6, the training process of the artificial neural network model is as follows:
adopting a supervised learning training mode, after a training set is input into an input layer, enabling each neuron to flow into a corresponding neuron in the next layer, summing and transmitting in a hidden layer, and finally, enabling the training set to reach an output layer for processing;
once the artificial neural network model calculates the output corresponding to one of the inputs, a loss function calculates an error vector, the loss function being a mean square error function: as shown in formula (I):
Figure BDA0003418007300000073
in formula (I), x is the input vector in the training set, y (x) is the output generated by the artificial neural network, y is the desired output, n is the size of the training set, w is the weight vector, b is the bias;
obtaining gradients of all parameters of the artificial neural network model by using a back propagation algorithm, and updating all parameters of the artificial neural network model by using a Levenberg-Marquardt optimization algorithm;
and finishing the training when the loss function converges to a certain degree or the iteration times are finished, and storing the parameters of the trained artificial neural network model to obtain the trained artificial neural network model. In the training stage, the number of nodes, weight vectors w and bias b of an input layer, an output layer and a hidden layer in the artificial neural network structure are modified through an optimal algorithm, so that a loss function is minimum.
Example 3
The GNSS signal capturing method based on the artificial neural network according to embodiment 1 or 2 is different in that:
processing digital intermediate frequency signals output by a radio frequency front end of a GNSS receiver, sequentially searching all PRNs by the GNSS receiver in a signal acquisition stage, storing the maximum value of correlation peak values of search frequency points, and generating a two-dimensional array data set A taking a frequency index value as an X axis and a correlation peak value corresponding to each frequency index value as a Y axiss。
Taking the application of the Beidou signal to the digital intermediate frequency signal processing part as an example, m is [1, 41 ]]，fMPIs 10KHz, fIFIs 0.098MHz, fbinAt 500Hz, t at 1023, tbinIs 1, fMCIs 100Hz, fcminIs 100Hz, o is [1, 3 ]]PRN is [1, 37 ]]. The method comprises the following steps:
step 1: setting the satellite number PRN which needs to be captured currently and belongs to PRN, and setting the frequency f generated by a local carrier generator to be fIF-fMP+(i-1)fbinI e m, the received digital intermediate frequency signal SIF(n) are mixed with the sine carrier signal I and the cosine carrier signal q generated by the local carrier generator respectively to obtain a baseband complex signal s (n) ═ I + jQ.
Step 2: and (4) carrying out Fourier transform on the baseband complex signal s (n) obtained in the step (1) to obtain a transform result X (k).
And step 3: the local pseudo-code generator generates a pseudo-code signal C (n) according to the current PRN to obtain C (k) through Fourier transformation, and C is obtained through conjugation*(k) Multiplying the result X (k) obtained in the step 2 to obtain Y (k).
And 4, step 4: and (5) performing inverse Fourier transform on the result Y (k) obtained in the step (3) to obtain y (n).
And 5: performing modular squaring on the result y (n) obtained in the step 4 to obtain a correlation peak value array A in the time domain, and solving the average value of the correlation peak value array A
Figure BDA0003418007300000081
Step 6: detecting the result of step 5 by using a mean value correlation peak value detection method, if
Figure BDA0003418007300000082
Consider that the current satellite PRN has been acquired, resulting in fC＝fIF-fMP+(i-1)fbinStep 7 is performed, otherwise step 8 is performed.
And 7: setting the frequency f-f generated by a local carrier generatorC-fMC+L/-1)fcminJ is equal to o, if j is less than o, j +1, repeating steps 1 to 5, storing the maximum value of the correlation peak value of each search frequency point, if j is equal to o, obtaining the correlation peak value of ApeakComposed search matrix AsStep 10 is performed.
And 8: if i < m, i +1, steps 1 to 6 are repeated, and if i ═ m, the current PRN satellite is considered invisible, step 9 is performed.
And step 9: and if PRN is less than PRN _ N, PRN +1 is carried out, and the steps 1 to 6 are repeated, and if PRN is equal to PRN _ N, the Beidou signal is captured completely.
Step 10: will matrix AsTraining and maximum value prediction are carried out by an artificial neural network model, and the structure of the artificial neural network model is according to the matrix AsIs preset.
Step 11: training a multi-layer perceptron neural network based on a Levenberg Marquardt learning algorithm, wherein the training cutoff condition is that iteration times are finished or errors meet requirements, and obtaining an optimal neural network structure and parameters according to a local optimal principle.
Step 12: and after training is finished, inputting test data to test the detection effect of the artificial neural model. And analyzing and checking the test data, including skewness and peak value, obtaining the spatial distribution of the peak value, and predicting the position pos of the maximum value.
Step 13: obtaining predicted high-precision carrier frequency f after Beidou signal acquisition is finishedPRN＝fC-fMC+(pos-1)fcmin。
In this embodiment, 7 test sets are used to compare the carrier frequency predicted based on the artificial neural network with the carrier frequency captured by the conventional capture algorithm, and table 1 lists the carrier frequency of the simulated intermediate frequency data, the carrier frequency predicted based on the artificial neural network, the error and the Matlab simulation time in this embodiment. Table 2 lists the carrier frequency of the simulated intermediate frequency data, the carrier frequency captured based on the conventional capture algorithm, the error, and the Matlab simulation time in this example.
TABLE 1
Figure BDA0003418007300000091
TABLE 2
Figure BDA0003418007300000092
As can be seen from table 1 and table 2, the GNSS signal capturing method based on the artificial neural network is superior to the carrier frequency captured based on the conventional capturing algorithm in terms of capturing accuracy and precision, is about one fifth of the carrier frequency captured based on the conventional capturing algorithm in terms of time, and greatly improves efficiency. The GNSS signal capturing method based on the artificial neural network has the advantages that the capturing precision and the capturing efficiency are considered simultaneously, and the GNSS signal capturing with high speed and high precision can be realized.
Example 4
A computer device comprising a memory storing a computer program and a processor implementing the steps of the artificial neural network-based GNSS signal acquisition method according to any one of embodiments 1 to 3 when the computer program is executed.
Example 5
A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the artificial neural network-based GNSS signal acquisition method according to any one of embodiments 1 to 3.

Claims (8)

1. A GNSS signal capturing method based on an artificial neural network is characterized by comprising the following concrete implementation steps:
step A: acquiring a data set comprising:
step 1: obtaining a correlation peak value in a time domain;
step 2: detecting the correlation peak value in the time domain obtained in the step 1 by adopting an average correlation peak detection method, and judging whether the current satellite is captured or not; the method comprises the following steps: if the ratio of the maximum correlation peak value to the average correlation peak value is larger than a set threshold value, entering the step 3; otherwise, entering step 4; wherein, the maximum correlation peak value refers to the maximum value in the correlation peak values in the time domain obtained in step 1, and the average correlation peak value is the average value of the correlation peak values in the time domain obtained in step 1;
and step 3: reducing the frequency searching step length, executing step 1, judging whether the carrier frequency searching is finished, and if the carrier frequency searching is finished, intercepting the maximum correlation peak value of the current frequency point and the maximum correlation peak value of the frequency points nearby as a data set; the data set comprises a training set and a test set; otherwise, continuing to execute the step 1;
and 4, step 4: judging whether the carrier frequency searching is finished or not, if so, determining that the current satellite PRN is not captured, adjusting the next satellite PRN, and performing the step 5, otherwise, performing the step 1;
and 5: judging whether the satellite PRN search is finished, if so, finishing the acquisition of the GNSS signal, otherwise, executing the step 1;
and B: training an artificial neural network model, comprising:
step 6: inputting the training set obtained in the step A into an artificial neural network model for training and maximum value prediction; the method specifically comprises the following steps: training an artificial neural network model, wherein the training cutoff condition is that iteration times are finished or errors meet requirements, and obtaining an optimal neural network structure and parameters, namely the trained artificial neural network model, according to a local optimal principle;
inputting test data in the test set into the trained artificial neural network model, testing the detection effect of the trained artificial neural network model, and predicting the position of the maximum value;
and C: GNSS signal capture is carried out through a trained artificial neural network model;
during testing, the related peak test set data generated during GNSS signal capturing is obtained through the steps 1-5, the test set data is input into the trained artificial neural network model, calculation and output are carried out in the artificial neural network model along the data flowing direction, one-time prediction is completed, and GNSS signal capturing is achieved.
2. The method for capturing GNSS signals based on artificial neural network according to claim 1, wherein the step 1 is implemented by:
step 1.1: mixing the received digital intermediate frequency signal with a copy sine carrier signal and a copy cosine carrier signal of a certain frequency respectively to obtain a baseband complex signal;
step 1.2: carrying out Fourier transform on the baseband complex signal obtained in the step 1.1;
step 1.3: carrying out Fourier transform on the local copy pseudo code, and multiplying the conjugate value after Fourier transform by the result obtained in the step 1.2;
step 1.4: performing inverse Fourier transform on the result obtained in the step 1.3;
step 1.5: and (4) carrying out modulus square on the result obtained in the step 1.4 to obtain a correlation peak value in a time domain.
3. The method for capturing GNSS signals based on an artificial neural network according to claim 1, wherein in step 2, an average correlation peak detection method is used to detect the correlation peak value in the time domain obtained in step 1, specifically:
maximum correlation peak value apeak＝max(A)；
Mean correlation peak
Figure FDA0003418007290000021
N represents the length of the sequence subjected to fourier transform; a is a two-dimensional array which is generated after modular squaring and takes a pseudo-random code phase index value as an X axis and takes a correlation peak value corresponding to each pseudo-random code index value as a Y axis;
Figure FDA0003418007290000022
where m denotes the size of the frequency index value, fMPRepresenting the maximum Doppler shift, fIFIndicating the nominal carrier frequency, f, of the digital intermediate frequency signalbinRepresenting a carrier frequency search step; n represents the magnitude of the pseudo-code phase index value, t represents the pseudo-code length, tcRepresenting the maximum pseudo-code phase shift, tbinSearching a step size for the pseudo code phase; o denotes the size of the frequency index value after reducing the frequency search step size, fCRepresenting the center carrier frequency, f, for artificial neural network model predictionMCMaximum search Range, f, representing Artificial neural network model predictioncminRepresenting the carrier frequency search step predicted by the artificial neural network model, and N representing the satellite PRN sequence to be acquired.
4. The method as claimed in claim 1, wherein in step 2, the maximum correlation peak of the current frequency point and the maximum correlation peak of the frequency points in the vicinity thereof are intercepted as data sets, specifically: the current frequency point is taken as the center, the left side and the right side are respectively extended by p frequency points, and the maximum correlation peak value A corresponding to each frequency point is obtainedpeakComposed search matrix AsAs a set of data, it is possible to,
Figure FDA0003418007290000023
floor is a floor rounding function.
5. The GNSS signal capturing method based on artificial neural network of claim 1, characterized in that the artificial neural network model comprises an input layer, a hidden layer and an output layer, the optimization algorithm uses Levenberg-Marquardt optimization algorithm, the hidden layer uses Sigmoid activation function, the output layer uses linear activation function.
6. The method as claimed in claim 5, wherein in step 6, the artificial neural network model is trained as follows:
adopting a supervised learning training mode, after a training set is input into an input layer, enabling each neuron to flow into a corresponding neuron in the next layer, summing and transmitting in a hidden layer, and finally, enabling the training set to reach an output layer for processing;
once the artificial neural network model calculates the output corresponding to one of the inputs, a loss function calculates an error vector, the loss function being a mean square error function: as shown in formula (I):
Figure FDA0003418007290000031
in formula (I), x is the input vector in the training set, y (x) is the output generated by the artificial neural network, y is the desired output, n is the size of the training set, w is the weight vector, and b is the offset;
obtaining gradients of all parameters of the artificial neural network model by using a back propagation algorithm, and updating all parameters of the artificial neural network model by using a Levenberg-Marquardt optimization algorithm;
and when the loss function is minimum or the iteration times are finished, finishing the training, and storing the parameters of the trained artificial neural network model to obtain the trained artificial neural network model.
7. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor when executing the computer program performs the steps of the artificial neural network-based GNSS signal acquisition method of any of claims 1 to 6.
8. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the artificial neural network-based GNSS signal acquisition method according to any one of claims 1 to 6.
CN202111553951.8A 2021-12-17 2021-12-17 GNSS signal capturing method based on artificial neural network Active CN114236577B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111553951.8A CN114236577B (en) 2021-12-17 2021-12-17 GNSS signal capturing method based on artificial neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111553951.8A CN114236577B (en) 2021-12-17 2021-12-17 GNSS signal capturing method based on artificial neural network

Publications (2)

Publication Number Publication Date
CN114236577A true CN114236577A (en) 2022-03-25
CN114236577B CN114236577B (en) 2022-08-23

Family

ID=80758345

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111553951.8A Active CN114236577B (en) 2021-12-17 2021-12-17 GNSS signal capturing method based on artificial neural network

Country Status (1)

Country Link
CN (1) CN114236577B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115508773A (en) * 2022-10-27 2022-12-23 中国电子科技集团公司信息科学研究院 Time difference method multi-station passive positioning method, system, electronic equipment and storage medium
CN115524662A (en) * 2022-10-27 2022-12-27 中国电子科技集团公司信息科学研究院 Direction finding time difference combined positioning method and system, electronic equipment and storage medium
CN116009034A (en) * 2022-12-21 2023-04-25 湖南工商大学 Satellite signal capturing method, baseband signal processing unit, receiver and medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CA2359662A1 (en) * 1999-02-01 2000-08-03 Snaptrack, Inc. Method and apparatus for measurement processing of satellite positioning system (sps) signals
CN109188470A (en) * 2018-09-11 2019-01-11 西安交通大学 A kind of GNSS cheating interference detection method based on convolutional neural networks
CN109782304A (en) * 2019-01-29 2019-05-21 西安交通大学 A kind of the GNSS cheating interference detection method and system of signal acquisition phase
CN113406674A (en) * 2021-05-08 2021-09-17 山东交通学院 Navigation satellite weak signal capturing method based on deep convolutional neural network
US20210325548A1 (en) * 2020-04-15 2021-10-21 Onenav, Inc. Matching for gnss signals

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CA2359662A1 (en) * 1999-02-01 2000-08-03 Snaptrack, Inc. Method and apparatus for measurement processing of satellite positioning system (sps) signals
CN109188470A (en) * 2018-09-11 2019-01-11 西安交通大学 A kind of GNSS cheating interference detection method based on convolutional neural networks
CN109782304A (en) * 2019-01-29 2019-05-21 西安交通大学 A kind of the GNSS cheating interference detection method and system of signal acquisition phase
US20210325548A1 (en) * 2020-04-15 2021-10-21 Onenav, Inc. Matching for gnss signals
CN113406674A (en) * 2021-05-08 2021-09-17 山东交通学院 Navigation satellite weak signal capturing method based on deep convolutional neural network

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
HAOQING LI ET AL.: ""Deep Learning of GNSS Signal Correlation"", 《ION GNSS+ 2020》 *
姜满: ""卫星导航多干扰源的识别和定位"", 《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》 *
王镇: ""GNSS导航基带信号捕获算法与电路关键技术研究"", 《中国优秀博硕士学位论文全文数据库(博士) 信息科技辑》 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115508773A (en) * 2022-10-27 2022-12-23 中国电子科技集团公司信息科学研究院 Time difference method multi-station passive positioning method, system, electronic equipment and storage medium
CN115524662A (en) * 2022-10-27 2022-12-27 中国电子科技集团公司信息科学研究院 Direction finding time difference combined positioning method and system, electronic equipment and storage medium
CN115524662B (en) * 2022-10-27 2023-09-19 中国电子科技集团公司信息科学研究院 Direction finding time difference joint positioning method, system, electronic equipment and storage medium
CN115508773B (en) * 2022-10-27 2023-09-19 中国电子科技集团公司信息科学研究院 Multi-station passive positioning method and system by time difference method, electronic equipment and storage medium
CN116009034A (en) * 2022-12-21 2023-04-25 湖南工商大学 Satellite signal capturing method, baseband signal processing unit, receiver and medium
CN116009034B (en) * 2022-12-21 2023-11-07 湖南工商大学 Satellite signal capturing method, baseband signal processing unit, receiver and medium

Also Published As

Publication number Publication date
CN114236577B (en) 2022-08-23

Similar Documents

Publication Publication Date Title
CN114236577B (en) GNSS signal capturing method based on artificial neural network
CN106855628B (en) Rapid acquisition and tracking system and method for high-dynamic satellite navigation signals
US10222479B2 (en) Psuedo maximum likelihood tracking for global navigation satellite systems
US20230168385A1 (en) System for determining a physical metric such as position
US6525687B2 (en) Location-determination method and apparatus
JP4869022B2 (en) Satellite signal tracking device and satellite signal receiver including the same
CN112327334B (en) GNSS long code signal capturing method and device assisted by low-orbit satellite
US10429515B2 (en) Method and apparatus for GNSS signal tracking
TW201445168A (en) A receiver and method for satellite positioning and speed measuring
KR101489729B1 (en) Methods and apparatuses for selectively validating satellite positioning system measurement information
US20150204981A1 (en) Signal processing method for ultra-fast acquisition and tracking of severely attenuated spread spectrum signals with doppler frequency and apparatus thereof
CN113406671A (en) GNSS forwarding type deception jamming detection method based on C/N0-MV
WO2011028928A1 (en) Energy domain based peak reconstruction methods and apparatuses
US7995682B2 (en) Method and apparatus for performing signal processing using historical correlation data
CN102508265A (en) Signal separation estimation theory-based satellite navigation signal multipath interference suppression method
CN115441900B (en) Information-assisted high-dynamic weak spread spectrum signal rapid capturing method and device
Zhu et al. Fast fine acquisition algorithm of GPS receiver aided by INS information
CN113671547B (en) Improved high dynamic capture method, device, equipment and storage medium
CN113514796B (en) Passive positioning method, system and medium
Li et al. Highly sensitive weak signal acquisition method for GPS/compass
CN107181708A (en) Frequency estimating methods and location receiver
CN112764063A (en) Method for realizing capture processing and receiver
Qian et al. Multi-channel signal parameters joint optimization for GNSS terminals
CN116520363B (en) Multi-phase arm code ring phase discrimination method
JP2823045B2 (en) Target motion analysis method

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant