Weak signal capturing method based on PMF-FFT

Abstract

The invention provides a weak signal capturing method based on PMF-FFT, which comprises the following steps: firstly, dividing an input signal SIF into an I path and a Q path, then respectively mixing the I path and Q path signals with a local carrier to zero intermediate frequency, then carrying out data grouping on the mixed signals, carrying out filtering processing on each group of signals through a segment matching filter, and finally, carrying out filtering processing on the signals through a PMF-FFT; and performing short-time correlation, zero padding and FFT (Fast Fourier Transform) operation on the signal after matched filtering, performing symbol flipping position detection on an FFT operation result, eliminating a data group with bit jump points in data, and finally performing coherent and incoherent operation on a result after the jump points are eliminated. According to the scheme of the invention, the method can better adapt to the signal capture in a strong noise environment, and remarkably improves the weak signal capture sensitivity performance of a navigation receiver.

Description

Technical Field

[0001] The present invention relates to the technical field of weak signal acquisition in the field of satellite navigation, and particularly relates to a weak signal acquisition method based on PMF-FFT. Background Art

[0002] In view of the fact that in a strong noise environment, the GNSS (Global Navigation Satellite System) signals received by a navigation receiver are very weak, how to improve the receiver sensitivity in a weak signal scenario and ensure the reliability of positioning services is the focus and hotspot of current research.

[0003] The PMF-FFT (Partial Matched Filtering - Fast Fourier Transform) algorithm can reduce the acquisition time and improve the acquisition efficiency, and is more suitable for weak signal acquisition under long integration times. Increasing the coherent integration time in the PMF-FFT algorithm can significantly increase the signal-to-noise ratio gain and improve the algorithm performance. However, coherent accumulation is easily affected by the navigation data message jumps, and in extreme cases, it will instead reduce the acquisition gain. In addition, this algorithm is also affected by scalloping loss, which affects the acquisition performance. Summary of the Invention

[0004] Aiming at the problem that signals are difficult to acquire by traditional navigation receivers in a strong noise environment, the present invention proposes a weak signal acquisition method based on PMF-FFT. The present invention divides the data into blocks, uses differential operations to detect the positions where the signal undergoes symbol inversion, thereby greatly increasing the integration time. Then, zero-padding is used to effectively suppress the possible scalloping loss of the acquisition algorithm. Finally, the acquisition sensitivity is further improved through coherent and non-coherent algorithms.

[0005] To achieve the above objectives, the technical solution of the present invention includes the following steps:

[0006] Step S1: Divide the input signal S IF into two paths, I and Q. Then, mix the I-path and Q-path signals with the local carrier to zero intermediate frequency respectively. Next, perform data grouping on the mixed signals. Finally, filter each group of signals through a segmented matched filter.

[0007] Step S2: Perform short-time correlation, zero-padding, and FFT operations on the signals after matched filtering.

[0008] Step S3: Detect the symbol flip positions according to the FFT operation results.

[0009] Step S4: Eliminate the data groups with bit jump points in the data.

[0010] Step S5: Perform coherent and non-coherent operations on the results after eliminating the jump points.

[0011] Furthermore, step S2 specifically includes:

[0012] Set the integration length of each group of signals to L sample points, then the duration t of short-time correlation p = Lt s , and the branch output signal of short-time coherence of the i-th group of data is expressed as:

[0013]

[0014] where, I i represents the real part of the signal, Q i represents the real part of the signal, t s is the sampling interval, k represents the sample point index in each group of signals, c(kt s ) is the ranging code generated locally, f IF represents the signal frequency, f dl is the locally estimated Doppler frequency, A represents the amplitude coefficient, Δω represents the carrier frequency difference, R(τ) is the short-time correlation function of the pseudo-code, is the phase difference, then, the correlation result output by the i-th PMF is:

[0015]

[0016] Then, after padding zeros to R i , perform P-point FFT calculation, and the normalized amplitude of the w-th point FFT output is expressed as:

[0017]

[0018] where, M represents the number of sampling points, f d represents the Doppler frequency offset.

[0019] Furthermore, step S3 specifically includes:

[0020] Perform differential operations on the N groups of FFT data results, multiply each group of data by the conjugate of the adjacent data behind it, and a total of N - 1 groups of differential operation results are obtained. Then, compare the results of each group of differential operations, select the maximum value of each group of differential results for comparison. The data group with the smallest maximum value of the differential results is the data group with bit transitions, that is, the position of symbol flipping. The differential operation is expressed as: where, F S represents the result of FFT operation, S represents the group number of data grouping, and * represents taking the conjugate.

[0021] Furthermore, step S5 specifically includes:

[0022] The coherent integration algorithm accumulates the results after removing the jump points from F(w), and its gain Z ccExpressed as: Wherein, F n is the result after removing the jump points from F(w), M' is the number of integration accumulations, and the total coherent integration time is also M'ms;

[0023] Non-coherent integration is to square and add the result after removing the jump points from F(w). The non-coherent integration gain Z nc Expressed as:

[0024] The present invention proposes a weak signal acquisition method based on PMF-FFT. This method divides the data into blocks, uses differential operations to detect the positions where the signal undergoes symbol inversion, thereby greatly increasing the integration time. By padding with zeros, it effectively suppresses the possible scalloping loss of the acquisition algorithm. Finally, combining coherent and non-coherent algorithms further improves the acquisition sensitivity. Compared with the existing signal acquisition algorithm, the semi-bit algorithm, the method of the present invention can obtain a longer coherent integration time when overcoming bit jumps. The present invention can better adapt to signal acquisition in a strong noise environment and significantly improves the weak signal acquisition sensitivity performance of the navigation receiver. Description of the Drawings

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0026] Figure 1 is a flowchart of a weak signal acquisition method based on PMF-FFT provided by an embodiment of the present invention;

[0027] Figure 2 is a schematic diagram of the influence of zero-padding on the PMF-FFT algorithm provided by an embodiment of the present invention; wherein, (a) is the three-dimensional acquisition graph without zero-padding, and (b) is the three-dimensional acquisition graph after zero-padding;

[0028] Figure 3 is a comparison graph of the acquisition of differential detection proposed by the present invention and the traditional semi-bit integration method provided by an embodiment of the present invention; wherein, (a) is the three-dimensional acquisition graph of the traditional semi-bit algorithm, and (b) is the three-dimensional acquisition graph of the improved algorithm of the present invention;

[0029] Figure 4 is a comparison graph of the detection probability of the solution of the present invention and other methods under different input signal-to-noise ratios provided by an embodiment of the present invention. Detailed Embodiments

[0030] 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.

[0031] The present invention proposes a weak signal acquisition method based on PMF-FFT. As Figure 1 shown, the method includes:

[0032] Step 1: Divide the input signal into two paths of I and Q, then mix the I-path and Q-path signals with the local carrier to zero intermediate frequency respectively, then perform data grouping on the mixed signals, and finally filter each group of signals through a segmented matched filter;

[0033] Divide the input signal S IF into two paths of I and Q, then mix them with the locally generated carrier signals respectively, then divide the mixed signals into N groups, and each group of signals passes through a segmented matched filter, and after partial matched filtering processing of each group of signals separately, perform analysis.

[0034] The input signal is an intermediate frequency signal obtained after being processed by the radio frequency front end of the receiver. The data grouping is performed for 1-bit data. In this embodiment, 20 ms of data is divided into 20 groups.

[0035] Step 2: Perform short-time correlation, zero-padding, and FFT operations on the signals after matched filtering

[0036] Set the integration length of each group of signals to L sample points, then the duration t of short-time correlation p = Lt s , and the branch output signal of short-time coherence of the i-th group of data is expressed as:

[0037]

[0038] Among them, I i represents the real part of the signal, Q i represents the real part of the signal, t s is the sampling interval, k represents the sample point index in each group of signals, c(kt s ) is the ranging code generated locally, f IF represents the signal frequency, f dl is the locally estimated Doppler frequency, A represents the amplitude coefficient, Δω represents the carrier frequency difference, R(τ) is the short-time correlation function of the pseudo code, is the phase difference, then, the correlation result output by the i-th PMF is:

[0039]

[0040] Then for R i After padding with zeros, perform the FFT calculation at point P. The normalized amplitude of the FFT output at the w-th point is expressed as:

[0041]

[0042] where M represents the number of sampling points, and f d represents the Doppler frequency offset.

[0043] Step 3: Perform symbol flip position detection

[0044] Perform a difference operation on the N groups of FFT data results. Multiply each group of data by the conjugate of the adjacent data behind it. A total of N - 1 groups of difference operation results are obtained. Then compare the results of each group of difference operations, and select the maximum value of each group of difference results for comparison. The data group with the smallest maximum value of the difference results is the data group with bit transitions, that is, the position of the symbol flip:

[0045]

[0046] where F S represents the FFT operation result, S represents the group number of the data grouping, and * represents taking the conjugate.

[0047] Step 4: Eliminate the data groups with bit transition points in the data

[0048] Delete the data groups where the bit transition points are located according to the difference results, so as to suppress the influence of the transition position on the integration matrix.

[0049] Step 5: Perform coherent and non - coherent operations

[0050] The coherent integration algorithm is to accumulate the result of F(w) after eliminating the transition points. Its gain Z cc is expressed as: where F n is the result of F(w) after eliminating the transition points, M′ is the number of integration accumulations, and the total coherent integration time is M′ ms;

[0051] The non - coherent integration is to square and then sum the result of F(w) after eliminating the transition points. The non - coherent integration gain Z nc is expressed as:

[0052] Compare the method of the present invention with the half - bit method through simulation. The results are as follows:

[0053] Take a 1 - ms signal, sampling frequency f s= 20.46 MHz, signal-to-noise ratio S / N = -23 dB, perform 3-fold zero-padding operation on the data, and the simulation results are as follows Figure 2 shown. (a) is the 3D acquisition graph without zero-padding, and (b) is the 3D acquisition graph after zero-padding. It can be seen from the figure that after zero-padding, the output correlation peak of 3D acquisition is more obvious, the spectral resolution is higher, and the acquisition effect is better.

[0054] Within a duration of 20 ms, the signal-to-noise ratio S / N = -34 dB, Figure 3 is the acquisition comparison between the half-bit method and the long coherent integration method based on differential detection of the method of the present invention. (a) is the 3D acquisition graph of the traditional half-bit algorithm, and (b) is the 3D acquisition graph of the improved algorithm of the present invention. It can be seen that the acquisition effect in (b) is better than that in (a), and it is also because a longer integration time can obtain a higher signal-to-noise ratio gain.

[0055] The input signal-to-noise ratio SNR increases from -44 dB to -32 dB. Comparing the half-bit algorithm, the improved PMF-FFT non-coherent integration algorithm and the improved PMF-FFT coherent integration algorithm proposed in this paper, the results are as follows Figure 4 shown. It can be seen from the figure that compared with the traditional half-bit integration method, the processing method based on detection and then coherence can have better acquisition sensitivity.

[0056] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A weak signal acquisition method based on PMF-FFT, characterized in that, The method includes: Step S1, input signal S IF is divided into two paths of I and Q. Then, the signals of the I path and the Q path are respectively mixed with the local carrier to zero intermediate frequency. Next, the mixed signals are grouped, and finally each group of signals is filtered through a segmented matching filter; Step S2: performing short-time correlation, zero-padding, and FFT operation on the signal after matched filtering; Step S3: detecting the symbol flip position according to the FFT operation result; Step S4: removing the data groups with bit transition points in the data; Step S5: performing coherent and non-coherent operations on the result after removing the transition points.

2. The method according to claim 1, wherein The step S2 further includes: Set the integration length of each group of signals to L sample points, then the duration t of short-time correlation p = Lt s , and the branch output signal of short-time coherence of the i-th group of data is expressed as: Among them, I i represents the real part of the signal, Q i represents the real part of the signal, t s is the sampling interval, k represents the sample point index in each group of signals, c(kt s ) is the ranging code generated locally, f IF represents the signal frequency, f dl is the Doppler frequency estimated locally, A represents the amplitude coefficient, Δω represents the carrier frequency difference, R(τ) is the short-time correlation function of the pseudo-code, is the phase difference. Then, the correlation result of the output of the i-th PMF is: Then for R i After padding with zeros, perform the FFT calculation at point P. The normalized amplitude of the FFT output at the w-th point is expressed as: where M represents the number of sampling points, and f d represents the Doppler frequency offset.

3. The method according to claim 1, wherein The step S3 further includes: performing a difference operation on N groups of FFT data results, multiplying each group of data by the conjugate of the adjacent data behind it, and obtaining a total of N-1 groups of difference operation results. Then, comparing the results of each group of difference operations, selecting the maximum value of each group of difference results for comparison. The data group with the smallest maximum difference result is the data group where bit flipping occurs, that is, the position of symbol flipping. The difference operation is expressed as: where F S represents the FFT operation result, S represents the group number of data grouping, and * represents taking the conjugate.

4. The method according to claim 1, wherein The step S5 further includes: The coherent integration algorithm accumulates the result after removing the jump points from F(w), and its gain Z cc is expressed as: where F n is the result after removing the jump points from F(w), M′ is the number of integrations and accumulations, and the total coherent integration time is also M′ ms; The non-coherent integration is to square and sum the result after removing the jump points from F(w), and the non-coherent integration gain Z nc is expressed as:

Images