BOC modulation satellite signal anti-short-time multipath unambiguous tracking method based on multilayer perceptron

Abstract

The invention discloses a BOC modulation satellite signal anti-short-time multipath unambiguous tracking method based on a multilayer perceptron. The method comprises the following steps: determining to receive BOC modulation satellite signals, and constructing a local triangular peak function; constructing a reconstruction correlation function based on unambiguous capture according to the received BOC modulation satellite signal; setting a local triangular peak function as a target vector to a multilayer perceptron, inputting a reconstruction correlation function based on non-fuzzy capture as input data to an input layer of the multilayer perceptron, and updating a weight loss based on Kalman filtering; after the weight loss of the multi-layer perceptron is updated, obtaining a weight vector, and weighting the weight vector to the reconstruction correlation function based on the unambiguous capture to obtain a final correlation function result; and the final correlation function result is used for delay locking of a tracking ring, so that the receiver maintains stable tracking of the direct signal. According to the invention, the influence of short-time multipath interference and BOC signal secondary peaks can be effectively suppressed, so that the tracking precision and stability are remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of anti-multipath ambiguity-free tracking in the field of satellite navigation, and particularly relates to a method for anti-short-term multipath ambiguity-free tracking of BOC-modulated satellite signals based on a multi-layer perceptron. Background Art

[0002] With the wide application of the Global Navigation Satellite System (GNSS), users have higher and higher requirements for the navigation and positioning accuracy. However, in practical applications, GNSS signals are easily affected by the multipath effect. Especially in complex environments such as urban canyons and near dense buildings, the multipath effect will significantly reduce the signal tracking accuracy, resulting in an increase in positioning errors.

[0003] BOC (Binary Offset Carrier) modulation is a modulation method widely used in modern GNSS signals (such as GPS L1C, Galileo E1 OS, etc.). By splitting the signal spectrum to both sides of the carrier frequency, BOC modulation improves the anti-interference ability and multipath resolution ability of the signal. However, the autocorrelation function of BOC signals has multiple side peaks, which may cause the traditional Delay Locked Loop (DLL) to lock the wrong peak during tracking. For BOC-modulated satellite signals, its correlation peak becomes narrower due to the increase in the modulation order, but at the same time, corresponding side peaks are generated near the main peak, which will bring corresponding ambiguity to acquisition and tracking. Therefore, an ambiguity-free acquisition and tracking method that retains the sharpness of the main peak and eliminates or suppresses the side peaks has become the research focus.

[0004] The ambiguity-free method for suppressing or eliminating side peaks is usually achieved through the combined operation of correlation functions. Therefore, it is more sensitive to the short-term multipath effect of the received signal. Therefore, it is necessary to perform further processing to resist short-term multipath, so as to improve the tracking performance and enhance the performance of the finally provided PNT service.

[0005] In recent years, deep learning technology has made remarkable progress in the field of signal processing. As a classic feedforward neural network, the multi-layer perceptron (MLP) has strong non-linear modeling ability and can solve classification and regression problems by learning complex signal features. Applying MLP to the anti-multipath tracking of BOC signals can effectively distinguish direct signals and multipath signals and identify the main peak of the autocorrelation function, thus achieving ambiguity-free tracking.

[0006] However, there is currently no publicly available technical solution to apply MLP to the anti-short-term multipath ambiguity-free tracking of BOC-modulated satellite signals. Therefore, developing a method for anti-short-term multipath ambiguity-free tracking of BOC-modulated satellite signals based on MLP has important theoretical significance and practical application value. Summary of the Invention

[0007] The object of the present invention is to provide a method for anti-short-term multipath ambiguity-free tracking of BOC modulated satellite signals based on a multi-layer perceptron. By online training the weights, the reconstructed correlation function of the ambiguity-free method is shaped to solve the technical problem that the traditional ambiguity-free tracking method of satellite BOC modulated signals in the prior art is sensitive to short-term multipath signals in a multipath navigation channel.

[0008] To solve the above technical problems, the specific technical solution of the present invention is as follows:

[0009] A method for anti-short-term multipath ambiguity-free tracking of BOC modulated satellite signals based on a multi-layer perceptron, the method comprising the following steps:

[0010] Step S1: Determine the received BOC modulated satellite signal and construct a local triangular peak function;

[0011] Step S2: Construct a reconstructed correlation function based on ambiguity-free acquisition according to the received BOC modulated satellite signal;

[0012] Step S3: Set the local triangular peak function as the target vector to the multi-layer perceptron, input the reconstructed correlation function based on ambiguity-free acquisition as the input data into the input layer of the multi-layer perceptron, and update the weight vector based on Kalman filtering;

[0013] Step S4: After the weight vector of the multi-layer perceptron is updated, obtain the weight vector, and weight the reconstructed correlation function based on ambiguity-free acquisition to obtain the final correlation function result;

[0014] Step S5: Use the final correlation function result in a delay-locked tracking loop to enable the receiver to maintain stable tracking of the direct signal.

[0015] Further, the local triangular peak function constructed in step S1 is expressed as follows:

[0016]

[0017] where is the width of a subcarrier chip, T c is the chip width corresponding to a pseudo-code, n is the position of the correlation function point, and M represents the modulation order of the BOC modulated satellite signal.

[0018] Further, step S2 includes the following steps:

[0019] Step S21: Intercept the start and end of the subcarrier corresponding to each chip to obtain two new local subcarriers;

[0020] Step S22: Modulate the local pseudo-code with the two new local subcarriers to obtain two new local signals;

[0021] Step S23: Perform correlation operations on the two new local signals and the received signal with the carrier stripped to obtain two correlation results;

[0022] Step S24: Perform a combined operation on the two correlation results to obtain a combined correlation function;

[0023] Step S25: After normalizing the combined correlation function, weight it to the original correlation result to obtain a reconstructed correlation function based on unambiguous acquisition.

[0024] Furthermore, the reconstructed correlation function based on unambiguous acquisition is expressed as follows:

[0025]

[0026] Among them, R AWCCF (n) represents the reconstructed correlation function based on unambiguous acquisition, ΔR(n) represents the combined correlation function, represents the normalization operation on the combined correlation function, R(n) represents the original correlation result, || represents taking the absolute value, and max represents taking the maximum value.

[0027] Compared with the prior art, the present invention has the following beneficial technical effects:

[0028] Compared with the traditional unambiguous acquisition method that is prone to tracking errors and ambiguity problems in a short-time multipath environment, the present invention introduces a multi-layer perceptron and Kalman filtering on this basis to further shape the reconstructed correlation function, effectively suppressing the influence of short-time multipath interference and the side peaks of the BOC signal, thereby significantly improving the tracking accuracy and stability. At the same time, it has the advantages of being easy to implement in engineering and applicable to complex navigation environments. Description of the Drawings

[0029] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments of the present invention. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0030] Figure 1 It is a flowchart of the BOC-modulated satellite signal anti-short-time multipath unambiguous tracking method based on a multi-layer perceptron of the present invention.

[0031] Figure 2 It is a comparison of the correlation function using AWCCF for unambiguous acquisition before and after being processed by a multi-layer perceptron in Embodiment 1 of the present invention.

[0032] Figure 3This is a comparison chart of the tracking errors of the AWCCF method processed by MLP in Embodiment 2 of the present invention and other methods with the change of multipath delay.

[0033] Figure 4 This is a comparison chart of the signal tracking errors of the AWCCF method processed by MLP in Embodiment 3 of the present invention and other methods under a changing navigation multipath channel. Specific implementation manners

[0034] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0035] As Figure 1 shown, the BOC modulation satellite signal anti-short-term multipath unambiguous tracking method based on a multi-layer perceptron proposed by the present invention includes the following steps:

[0036] Step S1: Determine the received BOC modulation satellite signal and construct a local triangular peak function.

[0037] The received BOC modulation satellite signal is represented by BOC(m,n), where m represents the subcarrier frequency and n represents the spreading code rate. The modulation order of the BOC modulation satellite signal is expressed as M = 2m / n. For the normalized correlation main peak of the BOC modulation satellite signal, its span can be described by the chip width. The local triangular peak function is expressed as follows:

[0038]

[0039] Among them, is the width of a subcarrier chip, T c is the chip width corresponding to a pseudo code, n is the position of the correlation function point, and n = 0 when the chips are completely aligned. It can be seen from the local triangular peak function that the width of the main peak becomes times that of the corresponding BPSK modulation signal.

[0040] Step S2: Construct a reconstructed correlation function based on unambiguous acquisition according to the received BOC modulation satellite signal.

[0041] The implementation manner of unambiguous acquisition is to eliminate or suppress the secondary peaks beside the correlation main peak of the received signal to achieve the purpose of eliminating ambiguity. Usually, the implementation of unambiguous acquisition is achieved through further processing of the local pseudo code. The final result is the result of the combined operation of the correlation function of the received signal and the local pseudo code. While suppressing the secondary peaks, it will be more sensitive to the influence brought by some short-term multipath signals.

[0042] The reconstruction correlation function based on unambiguous acquisition in the present invention is processed by an unambiguous method based on assignment adjustment of the correlation function. The unambiguous method based on assignment adjustment of the correlation function includes the following steps:

[0043] Step S21: Intercept the head and end of the subcarrier corresponding to each chip to obtain two new local subcarriers.

[0044] The two new local subcarriers are expressed as:

[0045]

[0046]

[0047] Among them, X1(n) represents the first local subcarrier, n represents the discrete index of the subcarrier, k0 represents the positioning on the index of the current carrier, and X2(n) represents the second local subcarrier.

[0048] Step S22: Modulate the local pseudo-code with the two new local subcarriers to obtain two new local signals.

[0049] The modulation formula is as follows:

[0050] H1(n) = C(n + τ)X1(n)

[0051] H2(n) = C(n + τ)X2(n)

[0052] Among them, H1(n) represents the first local signal, H2(n) represents the second local signal, and C(n + τ) represents the local pseudo-code with a delay of τ.

[0053] Step S23: Perform a correlation operation on the two new local signals and the received signal with the carrier removed to obtain two correlation results.

[0054] The correlation operation formula is as follows:

[0055]

[0056] Among them, R1(n) represents the first correlation result, S BOC represents the received BOC modulation satellite with the carrier removed, R2(n) represents the second correlation result, represents the convolution operation.

[0057] Step S24: Perform a combination operation on the two correlation results to obtain a combined correlation function.

[0058] The combination operation formula is as follows:

[0059] ΔR(n) = |λ1(n) + R2(n)| - |R1(n) - R2(n)|

[0060] Among them, ΔR(n) represents the combined correlation function, which is the correlation function used for subsequent tracking after non-ambiguity processing.

[0061] Step S25: After normalizing the combined correlation function, weight it to the original correlation result to obtain the reconstructed correlation function based on non-ambiguity acquisition.

[0062] The reconstructed correlation function based on non-ambiguity acquisition is expressed as follows:

[0063]

[0064] Among them, R AWCCF (n) represents the reconstructed correlation function based on non-ambiguity acquisition, represents the normalization operation of the combined correlation function, R(n) represents the original correlation result, || represents taking the absolute value, and max represents taking the maximum value.

[0065] The original correlation function R(n) can be expressed as:

[0066]

[0067] The combined correlation function can be expressed as:

[0068]

[0069] Among them, represents a triangular peak function with a center width of T s of, λ0(n) represents the function of noise. Usually, noise is not correlated. Therefore, after normalizing ΔR, weight it to the original correlation result. After weighting, a value close to the original main peak value can be obtained. However, due to the characteristics of some signals being amplified during the combination operation, especially the short-time multipath signals will have a greater impact. Therefore, the following further processing of this correlation function is required:

[0070] Step S3: Set the local triangular peak function as the target vector to the multi-layer perceptron, input the reconstructed correlation function based on non-ambiguity acquisition as the input data into the input layer of the multi-layer perceptron, and update the weight vector based on the Kalman filter.

[0071] The hidden layer of the multi-layer perceptron updates the weight vector according to the local triangular peak function and the correlation function reconstructed by the non-ambiguity acquisition method. The updated kernel is the Kalman filter method, which relies on the state equation and the observation equation to update the weight vector. The state equation and the observation equation are respectively expressed as:

[0072] W(k + 1) = W(k)

[0073] Y(k) = h[W(k) + X(k) + k] + V(k) = Yr r (k) + V(k)

[0074] Wherein, W(k + 1) represents the weight vector of the neural network at the (k + 1)-th iteration moment, W(k) represents the weight vector of the neural network at the k-th iteration moment, Y(k) represents the predicted output at the k-th iteration moment, X(k) represents the input signal at the k-th iteration moment, k represents the discrete iteration index, h represents the measurement matrix, V(k) represents the noise term at the k-th iteration moment, and Y r (k) represents the reference output at the k-th iteration moment.

[0075] Step S4: After the weight vector of the multi-layer perceptron is updated, the weight vector is obtained, and the weight vector is weighted to the reconstruction correlation function based on non-ambiguous acquisition to obtain the final correlation function result.

[0076] The weight vector obtained after the weight vector of the multi-layer perceptron is updated is denoted as ω(τ). By using the method of weighting with ω(τ) to process the reconstruction correlation function based on non-ambiguous acquisition, the final correlation function is obtained, and the result of the final correlation function is expressed as follows:

[0077] R FINAL (n) = R AWCCF (n) · ω(τ).

[0078] Step S5: The result of the final correlation function is used in the delay lock tracking loop to enable the receiver to maintain stable tracking of the direct signal.

[0079] Using the methods of steps S1 - S4 to process the leading and lagging branches in the delay lock tracking loop respectively to obtain their respective correlation results. According to the characteristics of the correlation function, when the correlation results of the leading branch and the lagging branch are the same, it can be considered that the phase is aligned at this time, and the maximum value of the correlation peak is on the prompt branch. Otherwise, adjust forward or backward according to the difference between the two to reduce the phase difference.

[0080] The phase discrimination output is obtained by the non-coherent leading minus lagging amplitude method phase discrimination formula, and the phase discrimination formula is as follows:

[0081]

[0082] Wherein, D EMLP represents the Early - Late phase discrimination output based on the multi-layer perceptron method, I E represents the in-phase correlation output of the leading branch, Q E represents the quadrature correlation output of the leading branch, I L represents the in-phase correlation output of the lagging branch, Q LRepresents the quadrature correlation output of the lag branch. For each correlation output

[0083]

[0084] where P is the intensity of the intermediate frequency signal, represents the carrier phase error existing in the tracking loop, τ represents the code phase error of tracking, d represents the unilateral width of the correlation peak, respectively represent the code chip phase differences of and when the correlation function values are obtained here by the method of performing steps S1 - S4 on the extracted I and Q signals of the early and lag branches, and R FINAL (n) is obtained. The chip delay distances of the three branches usually take a difference of half of the unilateral width of the correlation peak, so the length here is

[0085] The final phase discrimination result is obtained through the phase discrimination formula, and loop feedback is performed through the phase discrimination result to ensure that the receiver maintains stable tracking of the direct signal, thereby achieving the goal of anti-short-term multipath unambiguous tracking of the BOC modulated satellite signal based on the multi-layer perceptron.

[0086] To illustrate the effectiveness and superiority of the present invention, the present invention is compared with the traditional method and the unambiguous method through simulation with three embodiments below. The multipath navigation channel in the simulation is generated by fitting the simulation with reference to the ITU standard document.

[0087] Embodiment 1

[0088] Take the data component of the Beidou BOC signal under the multipath navigation channel for 1 ms, with the sampling frequency f s = 40.92 MHz, and perform unambiguous acquisition on the data using the AWCCF method, and the obtained correlation results are as shown by the curve before processing in Figure 2 and perform multi-layer perceptron processing on it, and the obtained correlation function curve is as shown by the annotation after processing. It can be found from Figure 2 that the output correlation peak after multi-layer perceptron processing weakens the time delay expansion brought by some multipath signals, and the sharpness of the main peak is guaranteed.

[0089] Embodiment 2

[0090] In the data component of the Beidou BOC signal under the 1 ms multipath navigation channel, in order to simplify the analysis, take a stronger reflected multipath signal, whose amplitude fades by 3 dB relative to the direct signal, and the time delay of this multipath signal gradually increases from 0, and obtain the multipath signal in the traditional tracking method, the AWCCF unambiguous method, and the AWCCF method with MLP processing proposed by the present invention, that is, AWCCF + MLP in the figure, and it can be seen from Figure 3From the tracking error curve in [reference], it can be seen that the processed method effectively eliminates the sensitivity of the ambiguity-free method to short-term multipath signals.

[0091] Embodiment 3: Take a multipath navigation channel model under a certain moving distance to test the signal. Track the signal using the traditional tracking method, various ambiguity-free methods, and the method of the present invention, and obtain the tracking error situation under the signal change. From Figure 4 it can be seen that the method using MLP+AWCCF effectively reduces the tracking error, which is the most stable and smallest among several methods, demonstrating the superiority and practicality of the present invention.

[0092] It can be understood that the present invention is described through some embodiments. Those skilled in the art know that without departing from the spirit and scope of the present invention, various changes or equivalent replacements can be made to these features and embodiments. Additionally, under the teaching of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application belong to the scope protected by the present invention.

Claims

1. A method for anti-short-term multipath unambiguous tracking of BOC-modulated satellite signals based on a multi-layer perceptron, characterized in that The method includes the following steps: Step S1: Determine to receive the BOC-modulated satellite signal and construct a local triangular peak function; Step S2: Construct a reconstructed correlation function based on unambiguous acquisition according to the received BOC-modulated satellite signal; Step S3: Set the local triangular peak function as the target vector to the multi-layer perceptron, input the reconstructed correlation function based on unambiguous acquisition as the input data into the input layer of the multi-layer perceptron, and update the weight vector based on Kalman filtering; Step S4: After the weight vector of the multi-layer perceptron is updated, obtain the weight vector, and weight it to the reconstructed correlation function based on unambiguous acquisition to obtain the final correlation function result; Step S5: Use the final correlation function result in the delay lock tracking loop to enable the receiver to maintain stable tracking of the direct signal.

2. The BOC modulation satellite signal anti-short-term multipath unambiguous tracking method based on a multi-layer perceptron according to claim 1, characterized in that, The local triangular peak function constructed in Step S1 is expressed as follows: Among them, is the width of a subcarrier chip, T c is the chip width corresponding to a pseudo-code, n is the position of the correlation function point, and M represents the modulation order of the BOC-modulated satellite signal.

3. The method for anti-short-term multipath unambiguous tracking of BOC modulated satellite signals based on a multi-layer perceptron according to claim 1, characterized in that, Step S2 includes the following steps: Step S21: Intercept the start and end of the subcarrier corresponding to each chip to obtain two new local subcarriers; Step S22: Modulate the local pseudo-code with the two new local subcarriers to obtain two new local signals; Step S23: Perform a correlation operation on the two new local signals and the received signal with the carrier stripped to obtain two correlation results; Step S24: Perform a combination operation on the two correlation results to obtain a combined correlation function; Step S25: After normalizing the combined correlation function, weight it to the original correlation result to obtain the reconstructed correlation function based on unambiguous acquisition.

4. The BOC modulation satellite signal anti-short-term multipath unambiguous tracking method based on a multi-layer perceptron according to claim 1, characterized in that, The reconstructed correlation function based on unambiguous acquisition is expressed as follows: where R AWCCF (n) represents the reconstruction-related function based on unambiguous acquisition, and ΔR(n) represents the combined correlation function, represents the normalization operation on the combined correlation function, R(n) represents the original correlation result, || represents taking the absolute value, and max represents taking the maximum value.

Images