CN105005057A - Beidou navigation system D1 navigation message capture method - Google Patents

Abstract

The invention discloses a Beidou navigation system D1 navigation message capture method. The method can adapt to all kinds of bit jump, increases the coherent integral gain, eliminates erroneous carrier frequency estimation caused by correlation peak splitting, and can be used to effectively and reliably capture a Beidou navigation system D1 navigation message. The method comprises the following steps: multiplying the signal sequence of a received D1 navigation message containing interference by a local carrier complex sequence and carrying out frequency mixing to realize carrier Doppler frequency offset stripping; then, expanding a local repetition code in two ways at the same time to obtain a local repetition code A and a local repetition code B; performing correlation operation on each of the two local repetition codes and the frequency mixing result; judging the results of the two correlation operations to obtain a carrier Doppler frequency offset estimated value fd and a ranging code phase delay estimated value Tau; and capturing the D1 navigation message.

Description

A kind of catching method of triones navigation system D1 navigation message

Technical field

The present invention relates to navigation signal detection technique field, be specifically related to a kind of catching method of triones navigation system D1 navigation message.

Background technology

Beidou satellite navigation and positioning system (BeiDou Navigation Satellite System, BDS) be Chinese independent research, the GPS (Global Position System) (Global Navigation Satellite System, GNSS) of independent operating.Beidou receiver is responsible for the signal receiving big-dipper satellite transmitting, carries out processing the information such as rear acquisition navigation message, realizes positioning function.Receiver realizes above function and has first needed catching satellite-signal, signal capture process is the three-dimensional search process of satellite number, carrier doppler frequency deviation and ranging code phase delay to received signal, when local reference sequences and the satellite-signal of reference numeral reach thick synchronous and frequency approximately equal, relevant peaks can be produced, as the foundation of acquisition success by related operation.The main method of satellite signal acquisition has the search of serial search, Parallel frequency domain, parallel code phase search and matched filtering etc., wherein apply Fast Fourier Transform (FFT) (Fast Fourier Transform, FFT) parallel code phase search method hardware utilization is high, acquisition speed is fast, is widely used in actual receiver.

The modulation of triones navigation system D1 navigation message has secondary coding, and the D1 navigation message being namely 50bps in speed raises Neumann-Hoffman (NH) code being shaped with 20 bits.Signal code speed after modulation is 1kbps, and code is wide is 1 millisecond, modulates with navigation information code and spread and disposal plus.Application NH code is improved positioning performance, increases cross-correlation gain and strengthen the advantages such as anti-arrowband interference, but wide identical with conventional coherent integration time owing to modulating rear signal code, and bit saltus step may occur any time in coherent integration, and occurrence frequency is frequent.Bit saltus step can cause the output relevant peaks of parallel code phase search to divide at frequency axis, and related gain reduces, and the Doppler led to errors estimates or catches unsuccessfully, is unfavorable for that receiver realizes quick position.

For solving the bit jump problem in triones navigation system D1 navigation message acquisition procedure, the method proposed at present has zero expansion method, two-step capturing method and NH code synchronization acquistion method etc.Zero expansion method can eliminate the relevant peaks division that bit saltus step causes, but algorithmic characteristic limits coherent integration gain, and when input signal carrier-to-noise ratio is lower, capture effect is undesirable.Two-step capturing method can alleviate bit saltus step to the impact of catching, but longer calculated amount consuming time is comparatively large, and capture effect is undesirable when carrier-to-noise ratio is lower.NH code synchronization acquistion method alleviates the impact of bit saltus step, can obtain larger gain, but calculated amount is large, is unfavorable for hardware implementing.

Summary of the invention

In view of this, the invention provides a kind of catching method of triones navigation system D1 navigation message, various bit saltus step situation can be adapted to, add coherent integration gain, and eliminate relevant peaks and divide the carrier frequency mistake that causes and estimate, can effectively, reliably catch triones navigation system D1 navigation message.

The catching method of triones navigation system D1 navigation message of the present invention, comprises the steps:

Step 1, ground receiver comprises the burst of the D1 navigation message of interference, sequence carries out zero padding to received signal, the Received signal strength sequence after zero padding is multiplied with the local carrier complex sequences obtained according to carrier doppler frequency deviation rough estimate evaluation and carries out mixing, realizes carrier doppler frequency deviation and peels off;

Step 2, this locality reproduction code is expanded according to two kinds of modes simultaneously, obtain local reproduction code A and local reproduction code B, wherein, local reproduction code A is made up of the same pseudo-random code in two cycles, local reproduction code B is made up of the pseudo-random code of one-period and the reverse pseudo-random code of one-period, and wherein oppositely pseudo-random code is obtained by the every negate of former pseudo-random code; Respectively zero padding is carried out to this locality reproduction code A and local reproduction code B;

Step 3, reappears code with this locality of two in step 2 respectively by the mixing results obtained in step 1 and carries out related operation;

Step 4, sends the two-way correlation result obtained in step 3 into decision device, obtains carrier doppler frequency deviation estimated value according to decision rule with ranging code phase delay estimated value wherein, decision rule is as follows:

M ₁and m ₂be respectively the maximal correlation peak amplitude that local reproduction code A and local reproduction code B obtains;

Work as m ₁>=m ₂time, namely maximal correlation peak amplitude reappears code A by this locality and is formed, and the correspondence position of local reproduction code A maximal correlation peak amplitude is (f _d1, τ ₁), work as τ ₁during≤N, (f _d1, τ ₁) be and catch result, work as τ ₁during > N, (f _d1, τ ₁+ N-L) be and catch result; Wherein, f _dfor carrier doppler frequency deviation, τ is ranging code phase delay, and subscript 1 represents the correlation parameter of local reproduction code A, and subscript 2 represents the correlation parameter of local reproduction code B, and N is the sampling number of Received signal strength sequence in the code cycle, and L is FFT computing length;

Work as m ₁<m ₂time, maximal correlation peak amplitude reappears code B by this locality and is formed, and the correspondence position of local reproduction code B maximal correlation peak amplitude is (f _d2, τ ₂), work as τ ₂during≤N, (f _d2, τ ₂) be and catch result, work as τ ₂during > N, (f _d2, τ ₂+ N-L) be and catch result.

Further, described zero padding method is: sequence or local reproduction code sequence carry out a small amount of zero padding to received signal, make the sequence length after zero padding be 2 integral number power, and length is the shortest.

Beneficial effect:

(1) the present invention is directed to existing GNSS signal catching method to be applied to exist when triones navigation system D1 navigation message is caught and to there is the problems such as relevant peaks decay, relevant peaks division, the large length consuming time of acquisition algorithm calculated amount, efficiency are not high as caught result, local code expanding method is adopted to eliminate the impact of bit saltus step, the problem that there is multiple bit saltus step situation in integral time is caused for after increase coherent integration time, this locality is reappeared code expand in two ways, to adapt to each bit saltus step situation, add coherent integration gain simultaneously; According to the sequence characteristic of two-way correlated results, design decision device, can obtain the carrier doppler frequency deviation estimated value needed for subsequent treatment and ranging code phase delay estimated value.The present invention can increase and catches gain, reduces hardware burden and increases efficiency of algorithm simultaneously, can adapt to various bit saltus step situation.

(2) the present invention adopts part zero padding method that sequence length is met to be the integral number power of 2, though partial sequence aliasing can be caused, but the relevant peaks division that may cause when there is a bit saltus step can be eliminated in coherent integration time further simultaneously, reduce computation complexity simultaneously, and then improve entire system performance.

(3) directly according to catching result, to obtain carrier doppler frequency deviation estimated value different with ranging code phase delay estimated value from conventional method, the present invention uses decision device to judge correlation result rationality according to each correlation peak location and amplitude, reduce false-alarm probability, improve and catch accuracy.

Accompanying drawing explanation

Fig. 1 is that local reproduction code expands method module map.

Fig. 2 is that local reproduction code expands the correlated results schematic diagram of method in each bit saltus step situation.

Fig. 3 is signal to noise ratio (S/N ratio) curve after the coherent integration of three kinds of different catching methods acquisitions.

Fig. 4 is two kinds of method receiver performance characteristics (Receiver Operating Characteristic, ROC) curve.

Fig. 5 is that two kinds of methods catch result statistic histogram.

Embodiment

To develop simultaneously embodiment below in conjunction with accompanying drawing, describe the present invention.

The invention provides a kind of catching method of triones navigation system D1 navigation message.

Beidou receiver receives the satellite-signal be submerged in additive white Gaussian noise by receiving antenna, through the down coversion of radio-frequency front-end, filter and amplification and analog-digital conversion (Analog-to-Digital Converter, ADC) process, do not consider the impact of filter loss and quantizing noise, suppose that additive white Gaussian noise sequence is independent uncorrelated, according to triones navigation system D1 navigation message generation rule and conventional Beidou receiver radio frequency part principle of work, can obtain the signal expression that baseband signal processing module receives is:

S [k] is pending burst, and A is satellite ranging signal amplitude (V), τ is ranging code phase-delay value (ms), T _sfor sampling period (ms), f _dfor the carrier frequency value (Hz) after down coversion, represent carrier wave intermediate frequency (Intermediate Frequency, IF) f _iFfor carrier doppler frequency deviation value f _dand, for carrier wave initial phase value, d [k-τ/T _s] represent numeric data code sequence, c _pRN[k-τ/T _s] represent pseudo-random code sequence, c _nH[k-τ/T _s] representing the NH code sequence participating in secondary modulation, n [k] represents noise sequence.

If do not consider the impact of bit saltus step, the production process of conventional parallel code phase search method relevant peaks is that input signal is multiplied with orthogonal recurring sequence with local homophase respectively, non-coherent integration is carried out after using circular correlation to carry out 1 millisecond of coherent integration, and then obtain each carrier doppler frequency deviation test value and relevant peaks peak value corresponding to ranging code phase delay test value, the estimated value of desired parameters is drawn according to peak-peak position.Coherent integration process expression formula is:

Wherein Y _iand Y _qbe respectively the coherent integration results of homophase and quadrature branch, for ranging code phase delay estimated value with the difference of actual value τ, for carrier doppler frequency deviation estimated value with actual value f _ddifference, R (Δ τ) is correlation result, and N is the sampling number of burst in 1 millisecond, n _iand n _qbe respectively the additive white Gaussian noise sequence of homophase and quadrature branch.Non-coherent integration procedure expression is:

$T=Y_I^2+Y_Q^2=\frac{A^2}{4}{R}^2\left(\mathrm{\&Delta;}\mathrm{\&tau;}\right){\mathrm{sinc}}^2\left({\mathrm{\&pi;f}}_d{\mathrm{NT}}_s\right)+n_I^2+n_Q^2---\left(3\right)$

Wherein T is the relevant peaks peak value eliminating the impact of carrier wave initial phase, judges estimates of parameters as operation result.Related operation can be realized by FFT and IFFT computing, and the two paths of signals of (2) formula can be expressed as:

Y＝IFFT{FFT(s[k]exp(-j2πf _DkT _s))(FFT(c _LOC[k])) ^*} (4)

Wherein subscript " ^*" representing that plural number gets conjugation, FFT and IFFT represents FFT computing and IFFT computing respectively.

When considering bit saltus step, for certain specific estimating carrier frequencies value, coherent integration results corresponding to its different ranging code phase delay estimated value is:

Wherein N _τfor the code phase jump position of actual signal, Δ f _dfor the error of estimating carrier frequencies value and actual value, α ₁and α ₂can be expressed as:

α ₁＝πΔf _d(N _τ-1)T _s

α ₂＝πΔf _d(N+N _τ-1)T _s

For assessing the performance of each catching method, definition signal to noise ratio snr is maximal correlation peak amplitude when only having signal to input square value and correlation peak when only having additive white noise to input square expectation value, wherein with for testing result:

$SNR=\frac{{\left\{\right|Y_S({\tilde{f}}_d,\widetilde{\mathrm{\&tau;}})\left|\right\}}^2}{E\left\{\right|Y_N({\hat{f}}_d,\widehat{\mathrm{\&tau;}}){|}^2\}}---\left(6\right)$

For solving that existing catching method affects by bit saltus step, gain is limited and the problem such as efficiency is not high, the present invention proposes local code expanding method to improve acquisition performance, the local code sequence that design two-way is different, uses base-2FFT to realize the associative operation with receiving sequence after a small amount of zero padding.

Wherein, for the multiple bit saltus step situation that may occur in coherent integration process, this locality reappears after code is expanded by two kinds of modes and reappears code participation related operation as new this locality by the present invention.The first local reproduction code is made up of the same pseudo-random code in two cycles, can be expressed as { 1,1}; Reproduction code in the second this locality is made up of the reverse pseudo-random code of the pseudo-random code of one-period and one-period, and wherein oppositely pseudo-random code is obtained by the every negate of former pseudo-random code, can be expressed as { 1 ,-1}.Then carry out related operation to two-way random code simultaneously, and adjudicated by decision device by operation result, the maximum gain position of acquisition is catches result.Wherein, if there is not bit saltus step in coherent integration time, then the first local code obtains maximum gain; If there is twice bit saltus step, then the second local code obtains maximum gain; There is a bit saltus step, then determine maximum gain acquisition pattern by jump position.Two kinds of local code parallel processings, all can obtain in the different bit saltus step situations in coherent integration time and correct catch result, too increase maximum gain amplitude simultaneously.

In order to use base-2 Fast Fourier Transform (FFT) (radix-2Fast Fourier Transform, radix-2FFT), usually need to carry out zero padding operation to sequence, by the integral number power of sequence zero padding to 2.Different from zero that local sequence is at least mended 1 pseudo-random code cycle by conventional method, the present invention adopts part zero padding method, namely only mend the integral number power that a small amount of zero makes sequence length meet to be 2, and the zero padding not carrying out 1 pseudo-random code Cycle Length is expanded.Though use part zero padding method can cause partial sequence aliasing, the relevant peaks division that may cause when there is a bit saltus step can be eliminated in coherent integration time simultaneously further, reduce computation complexity simultaneously, and then improve entire system performance.

The triones navigation system D1 navigation message trapping module that the present invention proposes as shown in Figure 1, concrete steps are: step 1, ground receiver comprises the burst of the D1 navigation message of interference, sequence carries out zero padding to received signal, Received signal strength sequence through a small amount of zero padding is multiplied with the local carrier complex sequences obtained according to carrier doppler frequency deviation rough estimate evaluation, realizes the thick stripping of carrier wave; Step 2, expands and a small amount of zero padding the local pseudo-random code of the satellite to be captured stored, two local code sequences of formation, i.e. local reproduction code A and local reproduction code B; Step 3, uses FFT and inverse fast Fourier transform (Inverse Fast Fourier Transform, IFFT) computing to realize the related operation of sequence in step 1 and step 2; Step 4, sends the two-way correlation result obtained in step 3 into decision device, obtains carrier doppler frequency deviation estimated value and ranging code phase delay estimated value according to decision rule.

Fig. 2 gives correlation result, and thick line represents the sequence level change caused due to NH coded modulation, and 1 to 2046 in square frame represents the ranging code sequence of D1 navigation message one-periods, and 0 in square frame is expressed as and realizes base-2FFT and the appropriate zero padding carried out.C [k] represents the input signal sequence after zero padding operation, c _lOC1[k] and c _lOC2[k] is newly-designed two-way local code sequence, h ₁[k] and h ₂[k] represents the result of local reproduction code A and local reproduction code B and input signal sequence related operation respectively.Wherein Fig. 2 (a) represents that 2 milliseconds of input signals do not have bit saltus step, and Fig. 2 (b) represents that 2 milliseconds of signals exist 1 bit saltus step, and Fig. 2 (c) represents that 2 milliseconds of signals exist 2 bit saltus steps.

When not considering noise, two-way coherent integration results can be expressed as:

Wherein i=1 represents local reproduction code A sequence, and i=2 represents local reproduction code B sequence, and L is FFT computing length, gets the integral number power of 2, is also the length of local reproduction code sequence.When coherent integration time is 2 milliseconds and code searching is spaced apart 0.5 chip, valid data length is that 8184, FFT computing length gets 8192, namely mends 80.Do not consider Δ f _dimpact, the peak-peak of relevant envelope by ranging code phase delay τ ∈ [0, N) and bit saltus step generation position N _τdetermine, wherein N _τvalue in 0, N-τ and 2N-τ.Maximal correlation peak amplitude F _maxbe expressed as:

Wherein p [k] represents the product of list entries and corresponding local code sequence.For different bit jump position, its maximum correlation peak amplitude expression is consistent, but the local code obtaining maximal value can change with the change of bit jump position and ranging code phase delay, and this is also the basic foundation that this method detects τ.

Non-coherent integration form is used to represent correlation result.Because a small amount of zero padding destroys the correlativity of sequence, in position (8) formula that maximal correlation peak amplitude occurs, shorter sequence cumulative sum is approximately zero.Therefore, non-coherent integration results expression formula becomes one from two, and whether bit saltus step occurs not affecting final detected non-coherent integration squared magnitude.Correlation result expression formula is:

$S=\left\{\begin{array}{c}\frac{A^2}{4}{\left\{\underset{k=0}{\overset{2N-\mathrm{\&tau;}-1}{\&Sigma;}}p\&lsqb;k\&rsqb;e^{j2{\mathrm{\&pi;\&Delta;f}}_d{\mathrm{kT}}_s}\right\}}^2,\mathrm{\&tau;}\&Element;\&lsqb;0,\frac{N}{2})\\ \frac{A^2}{4}{\left\{\underset{k=N-\mathrm{\&tau;}}{\overset{L-1}{\&Sigma;}}p\&lsqb;k\&rsqb;e^{j2{\mathrm{\&pi;\&Delta;f}}_d{\mathrm{kT}}_s}\right\}}^2,\mathrm{\&tau;}\&Element;\&lsqb;\frac{N}{2},N)\end{array}\right.---\left(9\right)$

Derive to (9) further, can obtain final correlation result expression formula is:

$S=\left\{\begin{array}{c}\frac{A^2}{4}{\left\{\frac{sin\&lsqb;{\mathrm{\&pi;\&Delta;f}}_d(2N-\mathrm{\&tau;})T_s\&rsqb;}{\sin \left({\mathrm{\&pi;f}}_d{T}_s\right)}\right\}}^2\mathrm{\&tau;}\&Element;\&lsqb;0,\frac{N}{2})\\ \frac{A^2}{4}{\left\{\frac{sin\&lsqb;{\mathrm{\&pi;\&Delta;f}}_d(N+\mathrm{\&tau;})T_s\&rsqb;}{\sin \left({\mathrm{\&pi;\&Delta;f}}_d{T}_s\right)}\right\}}^2\mathrm{\&tau;}\&Element;\&lsqb;\frac{N}{2},N)\end{array}\right.---\left(10\right)$

Adjudicate final correlation result, determine final to catch result, namely obtain the Doppler frequency deviation estimated value of input signal, ranging code phase delay estimated value, table 1 lists the decision rule of decision device.

Table 1 local code expanding method decision device decision rule

Wherein, m ₁and m ₂be respectively the maximal correlation peak amplitude that local reproduction code A and local reproduction code B obtains.Work as m ₁>=m ₂time, namely maximal correlation peak amplitude reappears code A by this locality and is formed, and the correspondence position of local reproduction code A maximal correlation peak amplitude is (f _d1, τ ₁), work as τ ₁during≤N, (f _d1, τ ₁) be and catch result, work as τ ₁during > N, (f _d1, τ ₁+ N-L) be and catch result.Work as m ₁<m ₂time, maximal correlation peak amplitude reappears code B by this locality and is formed, then local reproduction code B maximal correlation peak amplitude correspondence position is (f _d2, τ ₂), work as τ ₂during≤N, (f _d2, τ ₂) be and catch result, work as τ ₂during > N, (f _d2, τ ₂+ N-L) be and catch result.

Use the superiority of Matlab emulation and real data testing authentication this method respectively.In Matlab emulation, sample frequency f _sfor 16.368MHz, front end bandwidth B _iFfor 8.184MHz, IF-FRE is 4.092MHz, and each chip gathers 2 points and participates in computing, and namely step-length is 0.5 chip.

For embodying this method to the raising of signal to noise ratio (S/N ratio), the signal of different bit jump position is used to carry out Monte Carlo simulation.Choose five cycles representative data analysis, the NH code sequence of data sequence is 00010, and make 2 milliseconds of input signal optional positions within first three cycle start, namely the bit number of transitions of input signal is respectively 0 time, 1 time and 2 times.The probability that bit saltus step does not occur actual triones navigation system D1 navigation message 2 milliseconds of signals is 27.5%, and the probability that saltus step is 1 time is 45%, and the probability that saltus step is 2 times is 27.5%, therefore adopts this group detection sequence to have good representativeness.

Signal to noise ratio (S/N ratio) envelope diagram after the coherent integration that Fig. 3 is input signal carrier-to-noise ratio when being 38dB-Hz obtains according to formula (6) definition, ordinate is the signal to noise ratio (S/N ratio) in units of dB, horizontal ordinate is the signal start time in units of millisecond, i.e. 2 milliseconds of signals random selecting starting point in first 3 milliseconds of 00010 sequence.Wherein square curve 1 represents the signal to noise ratio (S/N ratio) curve using traditional catching method to obtain, and asteroid 2 represents the curve that use zero expanding method obtains, and circular curve 3 represents the curve that the method using the present invention to propose obtains.Result in Fig. 3 shows, traditional catching method shows when not having bit saltus step to occur well, but when bit saltus step occurs, particularly during saltus step 2 times, serious relevant peaks division causes error detection or undetected survey.Zero expanding method is not by the impact of bit saltus step, but coherent integration gain is restricted while gain stabilization.And the method that the present invention proposes does not exist relevant peaks division, better signal to noise ratio (S/N ratio) is had to show.

Although local code expanding method improves acquisition performance, add computation complexity to a certain extent.For the base-2FFT computing that length is L, wherein comprise 0.5Llog ₂l complex multiplication and Llog ₂l complex addition.Because traditional catching method affects seriously by bit saltus step, therefore be not further analyzed.After K non-coherent integration, the operation times of zero expansion method and local code expansion method is respectively m ₁and m ₂:

m ₁＝(7K-1)Lm ₀+KL+(3m ₀K+1.5)Llog ₂L

(11)

m ₂＝(9K+2 ^K+1-4)Lm ₀+KL+(4.5m ₀K+3)Llog ₂L

The zero expansion method that finds as calculated is carried out operation times needed for 5 non-coherent integrations and is carried out operation times needed for 3 non-coherent integrations a little more than the inventive method, is respectively 3.96 × 10 ⁸with 3.73 × 10 ⁸, can be used for algorithm performance and compare.

False-alarm probability P _fawith detection probability P _dfor weighing the important indicator of acquisition algorithm effect, during use classic method, expression formula is respectively:

$P_{fa}(V_t,1)=\exp (-\frac{V_t}{2{\mathrm{\&sigma;}}_n^2})---\left(12\right)$

$P_d(V_t,1)=Q_1(\frac{\sqrt{T_0}}{{\mathrm{\&sigma;}}_n},\frac{\sqrt{V_t}}{{\mathrm{\&sigma;}}_n})---\left(13\right)$

Wherein V _tfor decision threshold, m rank normalization Marcum Q equation expression formula is:

$Q_M(a,b)=\frac{1}{a^{M-1}}{\&Integral;}_b^{+\mathrm{\&infin;}}{x}^M\exp (-\frac{x^2+a^2}{2})I_{M-1}\left(ax\right)dx---\left(14\right)$

I _m-1x () is first kind M-1 rank Bessel's function.

For decision device of the present invention, the false-alarm probability of the inventive method reduces, and can be expressed as:

$P_{fa}(V_t,1)=\exp \&lsqb;-\frac{(2+h)V_t}{2{\mathrm{h\&sigma;}}_n^2}\&rsqb;---\left(15\right)$

Fig. 4 provides receiver performance characteristic (Receiver Operating Characteristic, ROC) curve assessment algorithm performance.Horizontal ordinate is false-alarm probability, and ordinate is corresponding detection probability, is 38dB-Hz, emulates when coherent integration time 2 milliseconds in input signal carrier-to-noise ratio.Wherein curve 1 carries out the result of 3 non-coherent integrations for new method, curve 2 carries out the result of 5 non-coherent integrations for new method, curve 3 is the result that zero expanding method carries out 3 non-coherent integrations, and curve 4 is the result that zero expanding method carries out 5 non-coherent integrations.Emphasis comparison curves 1 and curve 4, the two calculated amount is similar to, and the algorithm performance that the present invention proposes is better than zero expanding method.

Triones navigation system D1 navigation message measured data is used to verify the superiority of put forward the methods in the present invention below.Measured signal is obtained by Big Dipper digital medium-frequency signal sampling thief SIS600, and sampling thief design parameter is set to: sample frequency 10MHz, analog-to-digital converter resolution 4 bit, IF-FRE 2.5MHz, and sampling pattern is real sampling.The real satellite signal that antenna receives, through the process completing radio-frequency front-end of oversampler, becomes digital signal and is input in PC and carries out data processing and proof of algorithm.Without loss of generality, use Monte Carlo method, concrete method of testing is as follows: (1) obtains the signal data of more than 10 seconds from sampling thief, and stored in PC; (2) use Matlab to be divided into groups by signal data, often organizing duration is 2 milliseconds; (3) get continuous 10 milliseconds of data, use zero expanding method to carry out 5 non-coherent integrations, record testing result; (4) in (3), get front 6 milliseconds of number plays in 10 milliseconds of data, the method using the present invention to propose carries out 3 non-coherent integrations, record testing result; (5) Data duplication (3) of 10 seconds and (4) 1000 times are used, each carrier doppler frequency deviation estimated value and the ranging code phase delay estimated value detecting acquisition of record; (6) utilize the signal trace structure of 2 rank FLL+3 rank phase-locked loop structures, result will be caught and be used for signal trace, and obtain estimates of parameters accurately, as reference data.

Use said method to analyze one section of Received signal strength, do not consider that signal carrier-to-noise ratio obtains catching for 1000 times the Doppler frequency deviation value of signal used for-1470Hz when fluctuating, ranging code phase-delay value is 1921 chips, and carrier-to-noise ratio is 38dB-Hz.Fig. 5 provides the statistics with histogram result of catching result.Fig. 5 (a) and (c) respectively describe use zero and expand the histogram that method carries out Doppler frequency deviation estimated value that 5 non-coherent integrations obtain and ranging code phase delay estimated value; Fig. 5 (b) and (d) describe use local code to expand histogram that method carries out Doppler frequency deviation estimated value that 5 non-coherent integrations obtain and ranging code phase delay estimated value.Histogram ordinate representative data statistics number, wherein the horizontal ordinate of Fig. 5 (a) and (b) is the value estimating frequency deviation, and unit is Hz, and scope is ± 5000Hz; The horizontal ordinate of Fig. 5 (c) and (d) is the value of estimated code phase delay, and scope is 0 to 2045 chips.Comparison diagram 5 (a) and (b) can find out, the frequency deviation estimated value distribution that the local code expanding method using the present invention to propose obtains than zero expanding method obtain more accurately and more concentrate on exact value; Contrast 5 (c) and (d) also can draw, new method is also better than zero expanding method to the accuracy of estimation of code phase delay.Two method calculated amount are suitable, and real data test result proves that the local code expanding method that the present invention proposes has more advantage compared with zero expanding method further.

In sum, these are only preferred embodiment of the present invention, be not intended to limit protection scope of the present invention.Within the spirit and principles in the present invention all, any amendment done, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (2)

1. a catching method for triones navigation system D1 navigation message, is characterized in that, comprises the steps:

Step 1, ground receiver comprises the burst of the D1 navigation message of interference, sequence carries out zero padding to received signal, the Received signal strength sequence after zero padding is multiplied with the local carrier complex sequences obtained according to carrier doppler frequency deviation rough estimate evaluation and carries out mixing, realizes carrier doppler frequency deviation and peels off;

Step 2, this locality reproduction code is expanded according to two kinds of modes simultaneously, obtain local reproduction code A and local reproduction code B, wherein, local reproduction code A is made up of the same pseudo-random code in two cycles, local reproduction code B is made up of the pseudo-random code of one-period and the reverse pseudo-random code of one-period, and wherein oppositely pseudo-random code is obtained by the every negate of former pseudo-random code; Respectively zero padding is carried out to this locality reproduction code A and local reproduction code B;

Step 3, reappears code with this locality of two in step 2 respectively by the mixing results obtained in step 1 and carries out related operation;

Step 4, sends the two-way correlation result obtained in step 3 into decision device, obtains carrier doppler frequency deviation estimated value according to decision rule with ranging code phase delay estimated value wherein, decision rule is as follows:

M ₁and m ₂be respectively the maximal correlation peak amplitude that local reproduction code A and local reproduction code B obtains;

Work as m ₁>=m ₂time, namely maximal correlation peak amplitude reappears code A by this locality and is formed, and the correspondence position of local reproduction code A maximal correlation peak amplitude is (f _d1, τ ₁), work as τ ₁during≤N, (f _d1, τ ₁) be and catch result, work as τ ₁during > N, (f _d1, τ ₁+ N-L) be and catch result; Wherein, f _dfor carrier doppler frequency deviation, τ is ranging code phase delay, and subscript 1 represents the correlation parameter of local reproduction code A, and subscript 2 represents the correlation parameter of local reproduction code B, and N is the sampling number of Received signal strength sequence in the code cycle, and L is FFT computing length;

Work as m ₁<m ₂time, maximal correlation peak amplitude reappears code B by this locality and is formed, and the correspondence position of local reproduction code B maximal correlation peak amplitude is (f _d2, τ ₂), work as τ ₂during≤N, (f _d2, τ ₂) be and catch result, work as τ ₂during > N, (f _d2, τ ₂+ N-L) be and catch result.

2. the catching method of triones navigation system D1 navigation message as claimed in claim 1, it is characterized in that, described zero padding method is: sequence or local reproduction code sequence carry out a small amount of zero padding to received signal, make the sequence length after zero padding be 2 integral number power, and length is the shortest.