CN103645483B - Beidou signal capturing method in weak signal environment - Google Patents

Abstract

The invention relates to a Beidou signal capturing method in a weak signal environment. The method comprises the following steps: S1, arranging a down-conversion unit and carrying out down conversion processing on a received intermediate-frequency sampling signal; S2, carrying out NH code stripping on the received signal respectively, carrying out transformation into a frequency domain signal, carrying out multiplying with a multiplexing sum of a local code frequency domain value and then carrying out inverse transformation into a time domain; S3, carrying out conjugate multiplication on a coherent accumulation value at current time and a coherent accumulation value at previous time and carrying out summation; and S4, carrying out taylor series expansion on an amplitude difference value of two-side spectral lines of the peak value of the coherent integration result, deriving a quasi-linear relation of a frequency value and the amplitude difference value, and solving a frequency estimation value by using the linear relation. According to the invention, correlation calculation can be carried out on coherent values corresponding all code phases by one time; the signal to noise ratio is improved and the detection efficiency is improved; and the squaring loss of the non-coherent integration can be reduced. A problem of bit flipping caused by navigation data modulation can be solved, thereby improving the detection probability; the code phase and carrier wave frequency-offset estimation precision is enhanced; and the calculating speed and the calculating precision are guaranteed and the good stability and practicability are realized.

Description

Big Dipper signal acquisition methods under a kind of weak signal environment

Technical field

The invention belongs to navigation signal detection technique field, more particularly to Big Dipper signal acquisition methods, can be used for the Big Dipper weak Signal capture.

Background technology

GPS（Global Navigation Satellite System, GNSS）Signal is through straight The spread-spectrum signal of sequence spread spectrum modulation is connect, the capture of the satellite navigation signals to modulating through Direct-Spread is GNSS system needs The matter of utmost importance of solution.And when signal conditioning is undesirable, such as, indoors, under the conditions of forest and urban environment, block, multipath It is more serious with the phenomenon such as interference, energy have it is more weaken and decline, received signal to noise ratio has and greatly deteriorates, and common catches The method of obtaining will defy capture and trace into navigation satellite signal.Current serial/parallel algorithm, quick segmentation algorithm, FFT/IFFT The subject matter solved Deng acquisition algorithm is to reduce amount of calculation, improve acquisition speed.With the development of large scale integrated circuit, calculate Bottleneck overcome, and problem is concentrated under low signal-to-noise ratio in the capture of signal, i.e., survey after improving by suitable algorithm Signal to noise ratio, reaches the capture to Low SNR signal.GNSS is by lengthening coherent integration and incoherent integration times to improve letter Make an uproar and compare, so as to improve the sensitivity of signal detection, but, under the coherent integration gain that navigation signal bit reversal, frequency deviation cause Drop and code skew, the restriction of the Squared Error Loss of non-coherent integration and Complex Channel to coherent integration length, all can be to highly sensitive The design and performance of degree receiver detection algorithm has a huge impact.

In order to improve signal to noise ratio and then improve detection probability, overcome navigation data to modulate brought bit reversal and ask Topic, while reducing the Squared Error Loss of non-coherent integration, and improves frequency offset estimation accuracy, proposes that differential coherent accumulative combines frequency domain frequency The method of deviation estimation algorithm.The method using being realized in the environment of FPGA and DSP, with very strong practical significance.

The content of the invention

In order to solve the above problems, Big Dipper signal acquisition methods under a kind of weak signal environment of the invention, which includes,

S1 down-converter units, the if sampling signal to receiving carry out down-converted；

The signal for receiving is carried out peeling off NH codes by S2 respectively, and transforms to frequency domain, the volume of returning to work with local code frequency domain value After multiplication, another mistake transforms to time domain；

S3 is by the conjugate multiplication of the coherent accumulation values at current moment and the coherent accumulation values of previous moment and sues for peace；

S4 makees Taylor series expansion to the Magnitude Difference of coherent integration results peak value both sides spectral line, is derived frequency values and width The directrix sexual intercourse of degree difference, is solved using this linear relationship and obtains frequency estimation.

On the basis of above-mentioned technical proposal, step S1 includes：If sampling signal r (n) to receiving is carried out Down-converted：

Wherein s (n) is baseband signal, noises of the P (n) for intermediate-freuqncy signal r (n), f₁To receive the intermediate frequency carrier frequency of signal Rate, f_IFTo set IF-FRE, △ f are step-size in search, and i is searching times.

On the basis of above-mentioned technical proposal, step S2 includes：Value correspondence by L 1ms after related is added, i.e., complete Into the coherent integration of L ms data,

Then the coherent integration of multiple Lms is completed,

Wherein FFT () represents the number of winning the confidence fast Fourier transform, and IFFT () represents the fast Fourier transforma for seeking signal Change, conj () represents the complex conjugate for seeking signal, s_iN () is represented according to num_SKIPmaxBase band after delay adopts signal, NH_k(i) The NH codes sequence for determining is represented, c (n) represents locally generated ranging code.R_iN () represents certain 1ms coherent integration results, Y_i(n) table Show i-th Lms coherent integration results.

On the basis of above-mentioned technical proposal, the stripping NH codes include,

Wherein, wherein FFT () represents the number of winning the confidence fast Fourier transform, and IFFT () represents the fast Fourier for seeking signal Inverse transformation, conj () represent the complex conjugate for seeking signal, s_iN () represents that base band adopts signal, c (n) to represent locally generated range finding Code, num_SKIPRepresent from sampled data start delay sampling points, its scope control in 1ms sampling numbers, NH_kI () represents NH The one cycle of code moves to right the ranking results of k positions, R_iN () represents certain 1ms coherent integration results,Represent the num_SKIPThe 20ms coherent integration results of the ranking results of individual correspondence NH ring shift right k positions,

WhereinRepresent the original position of 1ms data, k_maxRepresent correspondence NH code sequence signs.

Estimate product of the corresponding navigation data bits value of current two neighboring coherent accumulation value for ± 1 value, Ran Houyu Difference accumulation carries out coherent accumulation after being multiplied, and formula is as follows：

Wherein, Y^* _i-1N () represents the conjugation of the coherent accumulation values of previous moment, Y_iN () is expressed as the relevant tired of current time It is value added, a_i-1Represent the product of the value (value is ± 1) of the corresponding navigation data bits of current two neighboring coherent accumulation value, Z (n) Represent differential coherent accumulative result.

On the basis of above-mentioned technical proposal, original frequency is compensated, repeat and be iterated estimation J time, finally Obtain High-precision carrier frequency and code phase.

Wherein,Intermediate variable estimated value is represented, D represents peak value left and right sides spectral line Magnitude Difference, and Q represents coherent accumulation Hop count, Q_IFFT points are represented, that is, has Q Lms coherent integration, points are done for Q_IFFT, G represents the peak value of FFT, α₁ Take 4/ π.f_iniFrequency estimation obtained by expression differential coherence, N_cFor spreading code Cycle Length,Represent that High-precision carrier is estimated Evaluation.

Compared with prior art, the present invention is due to enabling all codes of correlation computations of this method using frequency domain related algorithm The corresponding correlation of phase place.And using coherent accumulation and differential coherent accumulative combination algorithm, improve signal to noise ratio and then improve Detection probability, reduces the Squared Error Loss of non-coherent integration；

The present invention adopts differential coherent accumulative simultaneously, have estimated navigation data bits saltus step, overcomes navigation data modulation institute band The bit reversal problem come, improves detection probability；Using the linear offset estimation model launched based on Tayor, in low complexity Secondary capturing is carried out under degree, code phase and Nonlinear Transformation in Frequency Offset Estimation precision is effectively increased.Reality is carried out using the module of FPGA+DSP Now, arithmetic speed and precision are ensured, with very strong stability and practicality.

Description of the drawings

Fig. 1 is the system architecture diagram of the present invention；

Fig. 2 is the coherent integration Elementary Function block diagram in present system；

Fig. 3 is the differential coherence Elementary Function block diagram in present system；

Fig. 4 is the secondary fine capturing unit functional block diagram in present system；

Fig. 5 is the experimental situation Elementary Function block diagram in present system.

Specific embodiment

Refer to Fig. 1 to Fig. 5 to elaborate the present invention.

Refer to Fig. 1, under a kind of weak signal environment of the invention Big Dipper signal acquisition methods include with

Lower step：

S1 down-converter units, the if sampling signal to receiving carry out down-converted；

Step S1 includes：If sampling signal r (n) to receiving carries out down-converted：

Wherein s (n) is baseband signal, noises of the P (n) for intermediate-freuqncy signal r (n), f₁To receive the intermediate frequency carrier frequency of signal Rate, f_IFTo set IF-FRE, △ f are step-size in search, and i is searching times.

The signal for receiving is carried out peeling off NH codes by S2 respectively, and transforms to frequency domain, the volume of returning to work with local code frequency domain value After multiplication, another mistake transforms to time domain；

Wherein step S2 includes：Value correspondence by L 1ms after related is added, that is, complete the coherent integration of Lms data,

Then the coherent integration of multiple Lms is completed,

Wherein FFT () represents the number of winning the confidence fast Fourier transform, and IFFT () represents the fast Fourier transforma for seeking signal Change, conj () represents the complex conjugate for seeking signal, s_iN () is represented according to num_SKIPmaxBase band after delay adopts signal, NH_k(i) The NH codes sequence for determining is represented, c (n) represents locally generated ranging code.R_iN () represents certain 1ms coherent integration results, Y_i(n) table Show i-th Lms coherent integration results.

The stripping NH codes include,

Wherein, wherein FFT () represents the number of winning the confidence fast Fourier transform, and IFFT () represents the fast Fourier for seeking signal Inverse transformation, conj () represent the complex conjugate for seeking signal, s_iN () represents that base band adopts signal, c (n) to represent locally generated range finding Code, num_SKIPRepresent from sampled data start delay sampling points, its scope control in 1ms sampling numbers, NH_kI () represents NH The one cycle of code moves to right the ranking results of k positions, R_iN () represents certain 1ms coherent integration results,Represent the num_SKIPThe 20ms coherent integration results of the ranking results of individual correspondence NH ring shift right k positions,

Wherein num_SKIPmaxRepresent the original position of 1ms data, k_maxRepresent correspondence NH code sequence signs.

S3 is by the conjugate multiplication of the coherent accumulation values at current moment and the coherent accumulation values of previous moment and sues for peace；

Which includes, estimates the product of the value that the corresponding navigation data bits value of current two neighboring coherent accumulation value is ± 1, Then coherent accumulation is carried out after being multiplied with difference accumulation, formula is as follows：

Wherein, Y^* _i-1N () represents the conjugation of the coherent accumulation values of previous moment, Y_iN () is expressed as the relevant tired of current time It is value added, a_i-1Represent the product of the value (value is ± 1) of the corresponding navigation data bits of current two neighboring coherent accumulation value, Z (n) Represent differential coherent accumulative result.

S4 makees Taylor series expansion to the Magnitude Difference of coherent integration results peak value both sides spectral line, is derived frequency values and width The directrix sexual intercourse of degree difference, is solved using this linear relationship and obtains frequency estimation.

Which includes compensating original frequency, repeats and is iterated estimation J time, finally gives High-precision carrier frequency And code phase.

Wherein,Intermediate variable estimated value is represented, D represents peak value left and right sides spectral line Magnitude Difference, and Q represents coherent accumulation Hop count, Q_IFFT points are represented, that is, has Q Lms coherent integration, points are done for Q_IFFT, G represents the peak value of FFT, α₁ Take 4/ π.f_iniFrequency estimation obtained by expression differential coherence, N_cFor spreading code Cycle Length,Represent that High-precision carrier is estimated Evaluation.

The present invention is further described below in conjunction with the accompanying drawings.

1）Realize environment

Big Dipper signal is gathered indoors, is processed as primary signal.

2）With reference to Fig. 2, the sampled data of intermediate frequency is obtained through receiver, to set intermediate frequency f_IFAs benchmark, with △ f to search Suo Buchang, hunting zone are with f_IFAs ± the f of benchmark, then the search of 2f/ △ f secondary frequencies is carried out altogether, according to formula

Intermediate-freuqncy signal is dropped to fundamental frequency carries out coherent integration process.The method for being first according to FFT-IFFT is quick in FPGA The reception signal for completing every 1ms is related to local signal, takes 20ms data and completes the stripping of NH codes, then by the correlation knot of Lms Fruit correspondence is added, and obtains the coherent integration results that length is Lms.Acquired results are passed to into DSP by FIFO, is carried out in dsp Hereinafter process.

3）With reference to Fig. 3, the data to obtaining from upper One function unit carry out differential coherence process, and formula is as follows

Wherein, Y^* _i-1N () represents the conjugation of the coherent accumulation values of previous moment, Y_iN () is expressed as the relevant tired of current time It is value added, a_iRepresent the product of the value (value is ± 1) of the corresponding navigation data bits of current two neighboring coherent accumulation value, Z (n) tables Show differential coherent accumulative result.The principle being constantly present according to useful signal in peak value, by by the peak-to-peak of adjacent coherent integration Value differed from and mould be compared, the phase relation before and after judgement between adjacent coherent integration, estimate a_iValue, to avoid Caused the phenomenon cancelled out each other, the detection probability of the signal of increase by the saltus step of navigation data bits.Difference accumulation method is adopted simultaneously Reduce Square loss.First peak value and the second peak value of Z (n) are searched, and obtains peakedness ratio, if the ratio is more than thresholding λ, table It is bright to capture satellite and then carry out process below, otherwise, then show do not have when having traveled through all frequencies and being all not reaching to threshold value Capture, then change a satellite and captured.

4）With reference to Fig. 4, the code phase and carrier frequency of gps signal, wherein code-phase are estimated that through method above The precision of position is set according to Practical Calculation amount and computation complexity, and the estimated accuracy of carrier frequency is △ f, by acquired results, The compensation of code phase and frequency is carried out to if sampling signal, the coherent integration of Lms is then carried out, to Q Lms coherent integrations knot Fruit carries out Q_IThe FFT of point, by formula

Wherein,Intermediate variable estimated value is represented, D represents peak value left and right sides spectral line Magnitude Difference, and Q represents coherent accumulation Hop count, Q_IFFT points are represented, that is, has Q Lms coherent integration, points are done for Q_IFFT, G represents the peak value of FFT, α₁ Take 4/ π.f_iniFrequency estimation obtained by expression differential coherence, N_cFor spreading code Cycle Length,Represent that High-precision carrier is estimated Evaluation.Offset estimation is carried out, second compensation is carried out to the data after first compensation then, again according to the method with reference to Fig. 4 After carrying out J iteration, final capture data are transferred to into weak signal tracking link, process is tracked.

FPGA used in the present invention adds the flush bonding processor that DSP builds, specific design scheme reference Fig. 5, by day Line receives " Beidou II " navigation signal, and radiofrequency signal is down-converted to intermediate-freuqncy signal by downconverted module, then using ten tunnels Sampling channel signal is sampled, intermediate-freuqncy signal after sampling is transferred to into FPGA module, is completed from intermediate frequency in the module Fundamental frequency is dropped to, and NH codes are peeled off and related realization coherent integration carried out with local signal, and wherein USB, RS232 and Internet is used for Communicated with host computer.In the incoming DSP of coherent integration data that FPGA has been processed, differential coherence is completed by DSP Task and high accuracy Frequency Estimation, substantially increase arithmetic speed using two CSTR and reduce the execution time.DSP module also may be used Realized with the processor using other with similar functions, such as ARM；This area research worker can select to close according to physical condition Suitable device.

In sum, only the preferred embodiments of the invention, does not limit protection scope of the present invention with this, all according to the present invention Equivalence changes that the scope of the claims and description are made and modification, are all within the scope of patent of the present invention covers.

Claims (5)

1. Big Dipper signal acquisition methods under a kind of weak signal environment, it is characterised in that：Which includes,

Step S1 down-converter unit, the if sampling signal to receiving carry out down-converted；

The signal for receiving is carried out peeling off NH codes by step S2 respectively, and transforms to frequency-region signal, with answering for local code frequency domain value After conjugate multiplication, another mistake transforms to time domain, obtains coherent accumulation values；

Step S3 is by the conjugate multiplication of the coherent accumulation values at current moment and the coherent accumulation values of previous moment and sues for peace, its bag Include：Estimate the product of the value of the corresponding navigation data bits of current two neighboring coherent accumulation values, after being then multiplied with difference accumulation Coherent accumulation is carried out, its concrete formula is as follows：

$Z\left(n\right)=|\underset{i=2}{\overset{M}{\mathrm{\Σ}}}{a}_{i-1}*{Y^{*}}_{i-1}\left(n\right)Y_i\left(n\right)|$

Wherein, Y^* _i-1N () represents the conjugation of the coherent accumulation values of previous moment, Y_iN () is expressed as the coherent accumulation at current time Value, a_i-1Represent the product of the value of the corresponding navigation data bits of current two neighboring coherent accumulation values, the navigation data bits take Be worth differential coherent accumulative result is represented for ± 1, Z (n)；

Step S4 makees Taylor series expansion to the Magnitude Difference of differential coherent accumulative result Z (n) peak value both sides spectral line, is derived frequency Rate value and the directrix sexual intercourse of Magnitude Difference, are solved using directrix sexual intercourse and obtain frequency estimation；

Wherein, first peak value and the second peak value of Z (n) are searched, and obtains peakedness ratio, if the ratio is more than thresholding, show to catch Receive satellite and then carry out process below, otherwise, then show not capture when having traveled through all frequencies and being all not reaching to threshold value Arrive, then change a satellite and captured, Z (n) represents differential coherent accumulative result.

2. Big Dipper signal acquisition methods under a kind of weak signal environment as claimed in claim 1, it is characterised in that：Step S1 Including：If sampling signal r (n) to receiving carries out down-converted：

$\begin{array}{c}s\left(n\right)=r\left(n\right)*e^{-j2\mathrm{\π}(f_{IF}+i*\mathrm{\Δ}f)}\\ =\{h\left(l\right)*c\left(n\right)*e^{j*2{\mathrm{\πf}}_1}+P\left(n\right)\}*e^{-j2\mathrm{\π}(f_{IF}+i*\mathrm{\Δ}f)}\end{array}$

Wherein s (n) is baseband signal, noises of the P (n) for intermediate-freuqncy signal r (n), f₁To receive the intermediate frequency carrier frequency of signal, f_IF To set IF-FRE, Δ f is step-size in search, and i is searching times.

3. Big Dipper signal acquisition methods under a kind of weak signal environment as claimed in claim 1, it is characterised in that：Step S2 Including：Value correspondence by L 1ms after related is added, that is, complete the coherent integration of L ms data,

Then the coherent integration of multiple Lms is completed,

$\left\{\begin{array}{c}{R}_1\left(n\right)=IFFT\[FFT\left(s_1\right(n)*{\mathrm{NH}}_k(n\left)\right)*conj\left(FFT\right(e\left(n\right)\left)\right)\]\\ R_2\left(n\right)=IFFT\[FFT\left(s_2\right(n)*{\mathrm{NH}}_k(n\left)\right)*conj\left(FFT\right(e\left(n\right)\left)\right)\]\\ \mathrm{.......}\\ R_L\left(n\right)=IFFT\[FFT\left(s_L\right(n)*{\mathrm{NH}}_k(n\left)\right)*conj\left(FFT\right(e\left(n\right)\left)\right)\]\end{array}\right.$

$Y_i\left(n\right)=\underset{i=1}{\overset{L}{\mathrm{\Σ}}}{R}_i\left(n\right)$

Wherein FFT () represents the number of winning the confidence fast Fourier transform, and IFFT () represents the inverse fast Fourier transform for seeking signal, Conj () represents the complex conjugate for seeking signal, s_iN () is represented according to num_SKIPmaxBaseband signal after delay, NH_kN () represents NH codes One cycle move to right the ranking results of k positions, e (n) represents locally generated ranging code, R_iN () represents that certain 1ms coherent integration is tied Really, Y_iN () represents the accumulated value of i-th Lms coherent integration, num_SKIPmaxRepresent the original position of 1ms data.

4. Big Dipper signal acquisition methods under a kind of weak signal environment as claimed in claim 3, it is characterised in that：The stripping NH Code include,

$\left\{\begin{array}{c}{R}_1\left(n\right)=IFFT\[FFT(s_1\left(n+{\mathrm{num}}_{SKIP}\right)*N{H}_k\left(n\right))*conj\left(FFT\left(e\left(n\right)\right)\right)\]\\ R_2\left(n\right)=IFFT\[FFT(s_2\left(n+{\mathrm{num}}_{SKIP}\right)*N{H}_k\left(n\right))*conj\left(FFT\left(e\left(n\right)\right)\right)\]\\ \mathrm{.......}\\ R_{20}\left(n\right)=IFFT\[FFT(s_L\left(n+{\mathrm{num}}_{SKIP}\right)*N{H}_k\left(n\right))*conj\left(FFT\left(e\left(n\right)\right)\right)\]\end{array}\right.$

$Y_{\left({\mathrm{num}}_{SIKP,}k\right)}\left(n\right)=\underset{i=1}{\overset{20}{\mathrm{\Σ}}}{R}_i\left(n\right)$

Wherein, FFT () represents the number of winning the confidence fast Fourier transform, and IFFT () represents the inverse fast Fourier transform for seeking signal, Conj () represents the complex conjugate for seeking signal, s_iN () is represented according to num_SKIPmaxBaseband signal after delay, e (n) represent this real estate Raw ranging code, num_SKIPRepresent from sampled data start delay sampling points, its scope control in 1ms sampling numbers, NH_k N () represents that the one cycle of NH codes moves to right the ranking results of k positions, R_iN () represents certain 1ms coherent integration results,(n) Represent the n-th um_SKIPThe 20ms coherent integration results of the ranking results of individual correspondence NH ring shift right k positions,

$[{\mathrm{num}}_{\mathrm{SKIP}\max },k_{\max }]=\max \left(\right|Y_{({\mathrm{num}}_{\mathrm{SKIP}},k)}\left|\right)$

Wherein num_SKIPmaxRepresent the original position of 1ms data, k_maxRepresent correspondence NH code sequence signs.

5. Big Dipper signal acquisition methods under a kind of weak signal environment as claimed in claim 1, it is characterised in that：Step S4 Also include：Original frequency is compensated, is repeated and is iterated estimation J time, finally give High-precision carrier frequency and code-phase Position,

$\hat{x}=\frac{D}{{\mathrm{AQ\α}}_1}=\frac{\sin c(Qx/Q_I)}{{\mathrm{G\α}}_1}D$

$\hat{f}=f_{ini}+\underset{i=0}{\overset{J-1}{\mathrm{\Σ}}}\hat{x}/N_c{Q}_I$

Wherein,Intermediate variable estimated value is represented, D represents peak value left and right sides spectral line Magnitude Difference, and Q represents the section of coherent accumulation Number, Q_IFFT points are represented, that is, has Q Lms coherent integration, points are done for Q_IFFT, G represents the peak value of FFT, α₁Take 4/ π, f_iniFrequency estimation obtained by expression differential coherence, N_cFor spreading code Cycle Length,Represent that High-precision carrier is estimated Value.