Rapid calculation method for improving satellite navigation signal capture resolution

Abstract

The invention provides a quick calculation method for improving the capture resolution of a satellite navigation signal, which comprises the following steps of: setting down-conversion frequency and capture parameters through a parameter control module, calculating a frequency shift matrix, performing down-conversion on a received intermediate-frequency satellite signal to obtain a baseband signal for signal capture, and obtaining the capture resolution of the satellite navigation signal. According to the method, fast capture is achieved through a partial matched filtering-fast Fourier transform method, an FFT result is subjected to cyclic shift and then multiplied by a frequency shift matrix to achieve frequency shift, the capture resolution is improved, correlation peak search is conducted on a PMF-FFT result and a frequency shift result, and a final high-resolution capture result is obtained. Under a certain capture data length and FFT point number, the capture resolution is rapidly improved through matrix multiplication, and the capture performance is further improved.

Description

Technical Field

[0001] The present invention relates to the field of navigation data calculation, and in particular to a method for capturing global satellite navigation signals. Background Art

[0002] To use global satellite navigation (GNSS) signals for positioning, speed measurement and timing, the GNSS satellite signals must be captured first. The goal of GNSS satellite signal capture is to obtain the approximate carrier frequency and C / A code phase of the visible GNSS satellite signal to provide initial conditions for subsequent signal tracking. For each GPS satellite, a two-dimensional search needs to be completed: a search for the carrier frequency of the GNSS signal and a search for the C / A code phase. According to the capture strategy, capture can be divided into serial, parallel and hybrid types; according to the different processing domains, capture can be divided into two processing methods: time domain and frequency domain.

[0003] (1) FFT-based parallel code capture

[0004] After data preprocessing, the intermediate frequency input signal is mixed with the local carrier and converted to the frequency domain through FFT, and then multiplied with the FFT conjugate output result of the local pseudo code signal. After inverse fast Fourier transform (IFFT), the code phase capture output can be obtained.

[0005] (2) FFT-based parallel frequency capture

[0006] The input digital intermediate frequency signal is mixed with the sine and cosine carriers copied in the receiver, and then correlated with the locally copied pseudo code. The obtained correlation result is Fourier transformed to be converted into a frequency domain signal, which reflects the intensity of the correlation result at each frequency component. The peak position is the carrier frequency of the received signal.

[0007] (3) Serial code capture based on matched filter

[0008] A matched filter is a filter whose frequency response is consistent with the input signal spectrum. Its function is to correlate the local codes of different phases with the input data to obtain the correlation energy of different code phases. When the phase of the local code is consistent with the input data, the correlation energy is maximum.

[0009] The capture algorithm based on the partial matched filter (PMF) and FFT structure is a two-dimensional parallel search algorithm in time and frequency. It can overcome the shortcomings of slow serial search and parallel code phase search when the Doppler frequency deviation is large, and overcome the serious attenuation of correlation peak when only using the matched filter capture method. It can not only quickly complete the code phase search, but also accurately estimate the frequency deviation when the Doppler frequency deviation is large. It is widely used in GNSS receivers.

[0010] In order to achieve high sensitivity and high precision capture, the coherent integration time needs to be increased (the sensitivity and resolution can be theoretically increased by 3dB for each doubling of the integration time). The existing method has the problem of exponential increase in computing resources. At the same time, due to the influence of data flipping, the increase in coherent integration time is also subject to certain restrictions. The present invention can quickly increase the frequency resolution without changing the coherent integration time, thereby improving the capture accuracy. Summary of the invention

[0011] In order to overcome the deficiencies of the prior art, the present invention provides a fast calculation method for improving the capture resolution of satellite navigation signals.

[0012] The invention uses matrix operation to quickly improve the capture resolution based on the rapid capture achieved by partial matched filter-fast Fourier transform (PMF-FFT).

[0013] The technical solution process of the present invention is as follows Figure 1 As shown, it includes down-converting the satellite navigation intermediate frequency data; based on the partial matched filter-fast Fourier transform (PMF-FFT) to achieve fast capture, the frequency domain shift matrix is ​​calculated and stored according to the capture parameters (coherent integration time, FFT points, sampling rate, etc.); matrix multiplication is used to achieve fast frequency domain shifting and improve capture resolution.

[0014] The technical solution adopted by the present invention to solve the technical problem includes the following steps:

[0015] Step 1: Set the down-conversion frequency and capture parameters through the parameter control module;

[0016] Step 2: Down-convert the received intermediate frequency satellite signal into a baseband signal for signal capture;

[0017] Step 3: Use the partially matched filter-fast Fourier transform (PMF-FFT) method to achieve fast capture;

[0018] Step 4: Calculate the frequency shift matrix A according to the capture parameters and store it;

[0019] Step 5: Perform cyclic shift expansion on the FFT result in step 3 to form matrix B, and multiply matrix A by B to obtain the captured result after frequency expansion;

[0020] Step 6: Perform peak search on the capture result in step 5 to obtain the final high-resolution capture result.

[0021] In step 1, the parameter control module calculates the down-conversion frequency and capture parameters through the intermediate frequency signal sampling rate, code rate, and intermediate frequency, and the capture parameters include sampling rate, coherent integration time, FFT points, and PMF length; wherein the down-conversion frequency domain and the intermediate frequency are the same.

[0022] In step 2, the frequency shift function of the signal after matched filtering is as follows:

[0023]

[0024] Where fs is the sampling frequency, N is the number of FFT points, 0≤n<N, f shift is the frequency shift parameter.

[0025] Perform FFT transformation on f(n) and we get:

[0026] A=[F(1)F(2)…F(n)]

[0027] A is quantized and stored to obtain a frequency shifting matrix.

[0028] The specific steps of the partial matched filter-fast Fourier transform (PMF-FFT) method in step 4 are as follows:

[0029] The serial search of code phase is realized by using matched filtering. The matched filtering is a finite impulse response (FIR) filter with variable coefficients, including a shift register, a multiplier group and a multi-input adder. When the received signal sequence slides over the local sequence, a correlation result is generated at each moment. When the two sequences are aligned in phase, the correlation result will have a peak output.

[0030] FFT is used to realize parallel frequency search. The frequency resolution of N-point FFT is 1 / NTc, where Tc is the coherent integration time. Since the frequency of FFT is discontinuous, there is a fence effect, which can be reduced by tail zero padding.

[0031] In step 5, the FFT calculation result is as follows: n , 0≤n<N, perform cyclic shift to form matrix B i =[X (i+N-1)modN ,X (i+N)modN ,X (i+N+1)modN ,X (i+N+2)modN ,X i ] T ;

[0032] Frequency shifting is achieved through matrix multiplication:

[0033]

[0034] Perform a peak search.

[0035] The beneficial effect of the present invention is that the traditional method improves the capture resolution and sensitivity by increasing the coherent integration time. As the integration time increases, the computing resource index increases and is affected by the data bit flipping, which affects the satellite navigation signal capture performance. However, under the condition of a certain capture data length (the coherent integration time remains unchanged), the present invention realizes the rapid expansion of the frequency resolution through matrix multiplication operation, improves the capture sensitivity and is not affected by the data flipping, which can further improve the satellite navigation signal capture performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 The fast algorithm for improving the capture resolution of satellite navigation signals of the present invention is a schematic diagram.

[0037] Figure 2 Comparison of the capture results of the traditional PMF-FFT method and the method of the present invention, Figure 2 (a) is the capture result of the traditional PMF-FFT method. Figure 2 (b) is the capture result of the method of the present invention. DETAILED DESCRIPTION

[0038] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0039] Taking the B3I signal as an example, the parameters are set as follows: sampling rate 62MHz, spread spectrum code rate 10.23e6, intermediate frequency 21.52e6MHz, Doppler frequency 250Hz, and capture data length 2ms.

[0040] Step 1: Set the down-conversion frequency to 21.52e6MHz, capture parameters: PMF points to 2000, FFT points to 62.

[0041] Step 2: The frequency after matched filtering is fs = 31000 Hz, and the PMF-FFT capture resolution is 500 Hz. Taking one migration as an example, f shift Take 250Hz and calculate the frequency shift matrix as:

[0042] A0=[0,0,0,0,1]

[0043] A1=[6+85i,6+255i,6-255i,6-85i,0]

[0044] A0 is the initial frequency shift matrix, and A1 is the first frequency shift matrix;

[0045] Step 3: Down-convert the signal;

[0046] Step 4: Each time the local pseudo code slides, 62 PMF operation results can be calculated. The FFT operation of these 62 PMF operation results is used to obtain X i, where 0<=i<62;

[0047] Step 4: Circular shift the FFT result to get matrix B:

[0048]

[0049] Frequency shifting is achieved through the following matrix multiplication:

[0050]

[0051] Step 5: Perform peak search on Y to obtain the final high-resolution capture result, such as Figure 2 (b) as shown.

[0052] Compared with traditional methods Figure 2 Compared with (a), the present invention can increase the frequency resolution without increasing the coherent integration length, thereby improving the capture resolution and having certain advantages.

Claims

1. A fast calculation method for improving the capture resolution of satellite navigation signals, characterized in that The steps include: Step 1: Set the down-conversion frequency and capture parameters through the parameter control module; Step 2: Down-convert the received intermediate frequency satellite signal into a baseband signal for signal capture; Step 3: Use the partially matched filter-fast Fourier transform (PMF-FFT) method to achieve fast capture; Step 4: Calculate the frequency shift matrix A according to the capture parameters and store it; Step 5: Perform cyclic shift expansion on the FFT result in step 3 to form matrix B, and multiply matrix A by B to obtain the captured result after frequency expansion; Step 6: Perform peak search on the capture result in step 5 to obtain the final high-resolution capture result.

2. The fast calculation method for improving satellite navigation signal capture resolution according to claim 1, characterized in that: In step 1, the parameter control module calculates the down-conversion frequency and capture parameters through the intermediate frequency signal sampling rate, code rate, and intermediate frequency, and the capture parameters include sampling rate, coherent integration time, FFT points, and PMF length; wherein the down-conversion frequency domain and the intermediate frequency are the same.

3. The fast calculation method for improving satellite navigation signal capture resolution according to claim 1, characterized in that: In step 2, the frequency shift function of the signal after matched filtering is as follows: Where fs is the sampling frequency, N is the number of FFT points, 0≤n<N, f shift is the frequency shift parameter. Perform FFT transformation on f(n) and we get: A=[F(1) F(2)...F(n)] A is quantized and stored to obtain a frequency shifting matrix.

4. The fast calculation method for improving satellite navigation signal capture resolution according to claim 1, characterized in that: The specific steps of the partial matched filtering-fast Fourier transform method in step 4 are as follows: The serial search of code phase is realized by using matched filtering. The matched filter is a finite impulse response filter with variable coefficients, including a shift register, a multiplier group and a multi-input adder. When the received signal sequence slides over the local sequence, a correlation result is generated at each moment. When the two sequences are aligned in phase, the correlation result will have a peak output. FFT is used to realize parallel frequency search. The frequency resolution of N-point FFT is 1 / NTc, where Tc is the coherent integration time. Since the frequency of FFT is discontinuous, there is a fence effect, which can be reduced by tail zero padding.

5. The fast calculation method for improving satellite navigation signal capture resolution according to claim 1, characterized in that: In step 5, the FFT calculation result is as follows X n , 0≤n<N, perform cyclic shift to form matrix B i =[X (i+N-1)modN ,X (i+N)modN ,X (i+N+1)modN ,X (i+N+2)modN ,X i ] T ; Frequency shifting is achieved through matrix multiplication: Perform a peak search.

Images