CN104280750B - Long code capture systems based on partial matched filter fft algorithm and method - Google Patents

Abstract

The invention discloses a kind of long code capture systems based on partial matched filter fft algorithm and method, system includes: signal receives memory module, local pseudo-code memory module, partial matched filter module, single element accumulated value buffer module, fast Fourier transform FFT module, two code elements direct coherent accumulation module, two code elements process coherent accumulation modules, cellular non-coherent accumulation module, peak value comparison module and judging module.Method includes: obtain baseband signal；Carry out partial matched filter computing；Carry out fast Fourier transform FFT；Carry out the coherent accumulation of two code element different weights；Carry out cellular non-coherent accumulation；Relatively peak value；Judge whether more than decision threshold.The present invention improves the long code acquisition performance of satellite navigation signals under large doppler skew, low signal-to-noise ratio environment；Save the hardware resource of long Code acquisition.

Description

Long code capture systems based on partial matched filter fft algorithm and method

Technical field

The invention belongs to communication technical field, further relate to the one in technical field of satellite navigation based on part Join long code capture systems and the side of filtering fast Fourier transform (Fast Fourier Transformation FFT) algorithm Method.The present invention can be used for realizing the long Code acquisition of small-signal environment Satellite navigation signal, improves in small-signal environment The long code acquisition performance of satellite navigation signals.

Background technology

The most perfect along with navigation positioning system, the range of application of navigation positioning system is wider, but at some Special environment, the forest etc. blocked such as indoor, seabed, tunnel, city that high-rise building is intensive, trees, small-signal environment The long code of Satellite navigation signal correctly captures and faces a biggest difficult problem.In this context, how to improve in small-signal environment The long code acquisition performance of satellite navigation signals becomes a problem the most important.

The long code Trapped problems of satellite navigation signals actually obtains the carrier frequency of all visible satellite navigation signals The rough estimate value of rate and code phase.In small-signal environment, capture the long code of satellite navigation signals, lengthen satellite navigation letter Number coherent integration and the incoherent integration times of long code and local pseudo-code be two key components, but letter dynamic at height, low Ratio of making an uproar is under environment, and the Doppler frequency shift needing the satellite navigation signals of capture is big, signal to noise ratio is low, and coherent integration time is navigated Text bit width and the impact of change so that the coherent integration gain of acquisition is limited.Therefore, coherent integration time and non-is improved Coherent accumulation gain, the method improving long Code acquisition is to improve the long code acquisition performance of small-signal environment Satellite navigation signal Main solution.

The patent " a kind of people's code capture method and device " of Beijing BDStar Navigation Co., Ltd.'s application is (specially Profit application number 201310728686.1, publication number CN103698783A) propose a kind of partial matched filter fft algorithm people's code and catch Obtain method and device.Baseband signal is mainly carried out partial matched filter after local pseudo-code segmentation by the method for this patent application Processing, complete code phase search, then the result to partial matched filter carries out fast Fourier transform FFT, complete Becoming frequency search, finally the result to fast Fourier transform FFT carries out noncoherent accumulation, thus realizes final catching Obtaining, the deficiency that the method for this patent application publication exists is, in the case of Doppler frequency shift is relatively big, partial matched filter FFT calculates There is bigger associated loss in method, in the case of signal to noise ratio is low, partial matched filter fft algorithm acquisition performance is poor, Er Qieyun Calculation amount is big.The device of this patent application mainly includes the first filtering extraction module, the second filtering extraction module, partial matched filter Group, fast Fourier transform FFT processing module, non-coherent accumulation module and com-parison and analysis module, these modules are only limitted to portion Dividing the direct fast Fourier transform of data of matched filtering group, need substantial amounts of quick fourier transformation FFT hardware resource, this is special The deficiency that the disclosed device of profit application exists is that data are directly carried out after local pseudo-code partial matched filter soon by baseband signal Speed Fourier transformation FFT, thus the expense of hardware resource is big.

Chinese Academy of Sciences's Jiaxing microelectronics and the patent of system engineering center application " a kind of high sensitivity Beidou satellite navigation Signal acquisition methods and device " (number of patent application 201310637016.9, publication number CN103616702A) to propose one many The navigation signal method and device for capturing of phase coherence integration.The method of this patent application mainly carries navigation signal with local Ripple signal is divided into I road and Q road signal, then carries out monocycle integration, and multiple monocyclic integral results exist multiple phase-group Closing, multiple phase combination carries out coherent integration respectively, asks amplitude to be concerned with as a leggy integral result of same phase Multiple leggy integral results are carried out noncoherent accumulation, capture satellite navigation signals by integral result, this patent application publication The deficiency that exists of method be, although can capture satellite navigation signals, but, satellite navigation signals big at Doppler frequency shift In the case of weak, the method is difficult to capture satellite navigation signals.The device of this patent application mainly includes that local carrier occurs Device, local code generator, some multipliers, integrator and Acquisition Detection module, the function of these modules was only limitted to fixing week The navigation signal of phase and local pseudo-code coherent integration, it is impossible to the length of the dry integration that is concerned with according to the intensity adjustments of capacity of resisting disturbance, The deficiency that the device of this patent application publication exists is, the parameter of device cannot self-adaptative adjustment, capacity of resisting disturbance is low.

Patent " a kind of high sensitivity method for capturing satellite navigation signal and the system " (number of patent application of Ji'nan University's application 201210062429.4, publication number CN102608626A) propose a kind of coherence in frequency domain combine circular shifting search navigation letter Number catching method and system.Navigation signal after down coversion and sampling is mainly entered by the method for this patent application with this locality pseudo-code Row fast Fourier transform FFT, carries out satellite navigation signals and the local frequency domain correlation of pseudo-code, differential coherence integration and peak Value detection, and carry out frequency compensation by circular shifting search, detect and record peak-peak, the peak value obtained and detection threshold Relatively, thus realizing the correct capture of code phase and frequency, the deficiency that the method for this patent application publication exists is, circular shifting Operand big, under the conditions of maximum Doppler frequency shift, the success rate of signal capture is low.The system of this patent application includes that realization is defended Star digital medium-frequency signal Digital Down Convert operation Digital Down Converter Module, average sample satellite data and complete block accumulation function Average sample and block accumulator module, realize code phase frequency domain search FFT module profit, with Doppler's circular shifting search replace Frequency compensated circular shifting module is deposited, is stored up the local PRN code FFT of local PRN code FFT conjugation result and is conjugated memorizer, realization The complex multiplier module of signal despreading, calculate the IFFT module of different code phase correlated results, to the satellite-signal after despreading Carry out the cumulative differential coherence integration module of difference correlation energy, realize the peak detection block of signal capture output and to system The time-sequence control mode that each module sequential is controlled, the function of these modules is only limitted to navigation signal with local pseudo-code with linear Frequency domain correlation and linear differential coherence integration, it is impossible to realize the relevant matrix fortune with linear differential coherent integration of frequency domain Calculating, the deficiency that the system of this patent application publication exists is, satellite navigation signals and local pseudo-code are transported with linear frequency domain is relevant Calculate and linear differential coherence integration, cause satellite navigation signals slow in long Code acquisition data processing speed.

Summary of the invention

It is an object of the invention to overcome above-mentioned the deficiencies in the prior art, it is provided that a kind of based on partial matched filter FFT calculation The long code capture systems of method and method, realize maximum Doppler frequency shift, low noise with less hardware resource, faster processing speed Ratio long Code acquisition of navigation signal under environment, improves long code acquisition sensitivity and the capacity of resisting disturbance of satellite navigation signals, fall Low operand and the Doppler frequency shift produced under the high dynamic condition impact on acquisition performance can be overcome, it is thus achieved that be better than portion Divide the long Code acquisition of matched filtering fft algorithm.

The system of the present invention includes ten modules: signal receives memory module, local pseudo-code memory module, part coupling filter Mode block, single element accumulated value buffer module, fast Fourier transform FFT module, two code elements direct coherent accumulation module, two yards Unit processes coherent accumulation module, cellular non-coherent accumulation module, peak value comparison module and judging module；Wherein: signal receives and deposits Storage module, for being received the multiplier in memory module by signal, receives the satellite that in memory module, antenna receives by signal Navigation signal receives, with signal, the local oscillation signal that in memory module, voltage controlled oscillator produces and is multiplied, and obtains intermediate frequency satellite navigation letter Number, the intermediate frequency satellite navigation signals obtained is input to signal and receives in memory module in analog/digital A/D sampler；With Intermediate frequency satellite navigation signals is sampled by 4 times of bit rates of intermediate frequency satellite navigation signals, obtains digital intermediate frequency satellite navigation signals, will The digital intermediate frequency satellite navigation signals obtained is input to signal and receives band filter in memory module；By digital intermediate frequency satellite Navigation signal carries out baseband filtering, obtain with Doppler frequency deviation and white Gaussian noise baseband signal, the band that will be obtained The baseband signal having Doppler frequency deviation and white Gaussian noise is input in the sectionaliser 1 of partial matched filter module；Local pseudo-code Memory module, for being input to partial matched filter module by the local pseudo-code of the baseband signal of local pseudo-code memory module storage Sectionaliser 2 in；Partial matched filter module, for by the length of the baseband signal with Doppler frequency deviation and white Gaussian noise A is divided into B section, a length of C of subsegment of baseband signal, obtains subsegment baseband signal D, the subsegment baseband signal D input that will be obtained In correlator in partial matched filter module, wherein, the value of A is 204600, B=2^d, d is positive integer, and C is's Integer part；Use parallel search method, withFor step-length, the base that in selected part matched filtering module, sectionaliser 2 receives The code phase of the local pseudo-code of band signal, is divided into B section by length E of the local pseudo-code of baseband signal, and the subsegment of local pseudo-code is long Degree is F, obtains subsegment this locality pseudo-code G, subsegment this locality pseudo-code G obtained is input to the correlator of partial matched filter module In, wherein, the value of E is 204600, B=2^d, d is positive integer, and F isInteger part；By subsegment baseband signal D and son Local pseudo-code G of section carries out related calculation, and obtains single element coherent accumulation data stream；The single element coherent accumulation data stream that will be obtained It is input in single element accumulated value buffer module；Single element accumulated value buffer module, is used for utilizing table tennis random access memory The rate reduction of the single element coherent accumulation data stream that single element accumulated value buffer module is received by RAM, the list after being buffered Code element coherent accumulation data stream, the single element coherent accumulation data stream after the buffering that will be obtained is input to fast Fourier transform In FFT module；Fast Fourier transform FFT module, the solid size after the buffering that fast Fourier transform FFT module is received Unit's coherent accumulation data stream carries out L point quick Fourier conversion FFT, obtains frequency domain single element coherent accumulation data stream；To be obtained Frequency domain single element coherent accumulation data stream be input to two code elements direct coherent accumulation module and neutralize two code elements and process relevant tired In volume module；Two code elements direct coherent accumulation module, for being stored in M row g row by g frequency domain single element coherent accumulation data stream Adjacent two columns values in g row frequency domain single element coherent accumulation data stream are carried out respective value addition, obtain M row by matrixRow Two code elements direct coherent accumulation value, the M row that will be obtainedIt is incoherent that the two code elements direct coherent accumulation value of row is input to cellular In accumulation module, wherein, g is the even number more than or equal to 2, M=2^d, d is integer；Two code elements process coherent accumulation module, are used for G frequency domain single element coherent accumulation data stream is stored in the matrix of M row g row, in g row frequency domain single element coherent accumulation data stream The numerical value of even column negates, in g row frequency domain single element coherent accumulation data stream of the inverted for even column numerical value adjacent two Columns value carries out respective value addition, obtains M rowTwo code elements of row process coherent accumulation value, the M row that will be obtainedThe two of row Code element processes coherent accumulation value and is input in cellular non-coherent accumulation module；Cellular non-coherent accumulation module, for by M rowRow Two code elements direct coherent accumulation value and M rowTwo code elements process coherent accumulation values of row carry out respective column data respectively and take definitely Add up after value, obtain the direct cellular non-coherent accumulation value of M row 1 row and the process cellular non-coherent accumulation value of M row 1 row, by institute The direct cellular non-coherent accumulation value of M row 1 row obtained is input to peakedness ratio with the process cellular non-coherent accumulation value of M row 1 row In the peak value screening washer of relatively module；Peak value comparison module, arranges with M row 1 for the direct cellular non-coherent accumulation value arranged by M row 1 Process cellular non-coherent accumulation value take the maximum in M row 1 columns value respectively, obtain direct cellular non-coherent accumulation peak value with Process cellular non-coherent accumulation peak value, direct cellular non-coherent accumulation peak value and the process cellular non-coherent accumulation peak that will be obtained In peak-peak comparator in value input peak value comparison module；By direct cellular non-coherent accumulation peak value and the process non-phase of cellular Dry accumulation peak compares, and chooses direct cellular non-coherent accumulation peak value and processes maximum in cellular non-coherent accumulation peak value one Peak value, obtains the peak-peak of cellular non-coherent accumulation, is input to sentence by the peak-peak of the cellular non-coherent accumulation obtained Certainly in module；Judging module, for the peak-peak of cellular non-coherent accumulation is compared with decision threshold, if no more than decision gate Limit, then the code phase changing the local pseudo-code of search re-searches for；If more than decision threshold, then complete capture.

To achieve these goals, the thinking of the inventive method is: carry out list first with partial matched filter fft algorithm The partial matched filter of code element, then realizes the different weights of multiple-symbol segmentation coherent accumulation value with multi-code coherent accumulation algorithm Cumulative, thus increase the time of coherent accumulation Baud Length and coherent accumulation, finally utilize multi cell non-coherent accumulation algorithm real The non-coherent accumulation of existing multi cell, thus increase the number of times of non-coherent accumulation, it is achieved under maximum Doppler frequency shift, low signal-to-noise ratio environment The long Code acquisition of navigation signal.

The inventive method to realize step as follows:

(1) baseband signal is obtained:

(1a) receive the multiplier in memory module by signal, signal is received the satellite that in memory module, antenna receives Navigation signal receives, with signal, the local oscillation signal that in memory module, voltage controlled oscillator produces and is multiplied, and obtains intermediate frequency satellite navigation letter Number, the intermediate frequency satellite navigation signals obtained is input to signal and receives in memory module in analog/digital A/D sampler；

(1b) with 4 times of bit rates of intermediate frequency satellite navigation signals, intermediate frequency satellite navigation signals is sampled, obtain digital intermediate frequency Satellite navigation signals, is input to the digital intermediate frequency satellite navigation signals obtained signal and receives bandpass filtering in memory module Device；

(1c) digital intermediate frequency satellite navigation signals is carried out baseband filtering, obtain with Doppler frequency deviation and white Gaussian noise Baseband signal, the baseband signal with Doppler frequency deviation and white Gaussian noise obtained is input to partial matched filter In the sectionaliser 1 of module, baseband signal is expressed as:

R (n)=Ad (n) p (n+ τ) e^j2πfnΔT+N

Wherein, r (n) represents that n represents the sampling instant of baseband signal, and A represents in the baseband signal that sampling instant n receives The amplitude of baseband signal, d (n) represents the modulated symbol in sampling instant n baseband signal, and p (n+ τ) represents at sampling instant n base The long pseudo-random code of band signal, τ represents the offset phase of the long pseudo-random code of baseband signal, and e represents that natural constant, j represent empty Number unit, π represents that pi, f represent that baseband signal Doppler frequency deviation, Δ T represent the sampling time interval of baseband signal, N table Show the white Gaussian noise of baseband signal；

(2) partial matched filter computing is carried out:

(2a) length A of the baseband signal with Doppler frequency deviation and white Gaussian noise is divided into B section, the son of baseband signal Segment length is C, obtains subsegment baseband signal D, subsegment baseband signal D obtained is input in partial matched filter module In correlator, wherein, the value of A is 204600, B=2^d, d is positive integer, and C isInteger part；

(2b) the local pseudo-code of the baseband signal of local pseudo-code memory module storage is input to partial matched filter module In middle sectionaliser 2；

(2c) parallel search method is used, withFor step-length, in selected part matched filtering module, sectionaliser 2 receives The code phase of the local pseudo-code of baseband signal, is divided into B section, the subsegment of local pseudo-code by length E of the local pseudo-code of baseband signal A length of F, obtains subsegment this locality pseudo-code G, and subsegment this locality pseudo-code G obtained is input to the relevant of partial matched filter module In device, wherein, the value of E is 204600, B=2^d, d is positive integer, and F isInteger part；

(2d) utilize following formula, subsegment baseband signal D is carried out related calculation with subsegment this locality pseudo-code G, obtain single element and be concerned with Accumulation data stream；

$s\left(m\right)=\underset{n=\mathrm{mC}}{\overset{(m+1)C-1}{\mathrm{\Σ}}}r\left(n\right)q(n+\mathrm{\α})$

Wherein, single element coherent accumulation data when s (m) represents section serial number m, m represents single element coherent accumulation data Section sequence number, wherein m=0,1 ..., B-1, B represent the single element segments of baseband signal, and ∑ represents that sum operation, C represent base The subsegment length of band signal, r (n) represents that n represents the sampling instant of baseband signal, q in the baseband signal that sampling instant n receives (n+ α) represents the local pseudo-code in sampling instant n baseband signal, and α represents the local pseudo-code initial ranging code phase of baseband signal；

(2e) the single element coherent accumulation data stream obtained is input in single element accumulated value buffer module；

(3) fast Fourier transform FFT is carried out:

(3a) single element utilizing table tennis random access memory ram single element accumulated value buffer module to be received is relevant tired The rate reduction of volume data stream, the single element coherent accumulation data stream after being buffered, the single element after the buffering that will be obtained Coherent accumulation data stream is input in fast Fourier transform FFT module；

(3b) following formula is utilized, the single element coherent accumulation data after the buffering receive fast Fourier transform FFT module Stream carries out L point quick Fourier conversion FFT, obtains frequency domain single element coherent accumulation data stream；

$S\left(k\right)=S_1\left(k\right)+e^{\frac{-j2\mathrm{\πk}}{L}}{S}_2\left(k\right)$

$S(k+\frac{L}{2})=S_1\left(k\right)-e^{\frac{-j2\mathrm{\πk}}{L}}{S}_2\left(k\right)$

Wherein, before S (k) represents at half frequency domain sample point kThe frequency spectrum of item single element coherent accumulation data stream, k The half frequency domain sample point of expression single element coherent accumulation data stream, wherein,S₁K () represents at half frequency The frequency spectrum of the even number section sequence number single element coherent accumulation data stream at the sampled point k of territory, e represents that natural constant, j represent imaginary number list Position, L represents counting of fast Fourier transform FFT, wherein L=2^d, d is positive integer, S₂K () represents at half frequency domain sample point k The frequency spectrum of the odd number section sequence number single element coherent accumulation data stream at place,Represent at half frequency domain sample pointPlace AfterThe frequency spectrum of item single element coherent accumulation data stream；

(3c) the frequency domain single element coherent accumulation data stream obtained is input in two code elements direct coherent accumulation module Process in coherent accumulation module with two code elements；

(4) coherent accumulation of two code element different weights is carried out:

(4a) g frequency domain single element coherent accumulation data stream is stored in the matrix of M row g row, g row frequency domain single element is concerned with In accumulation data stream, adjacent two columns values carry out respective value addition, obtain M rowThe two code elements direct coherent accumulation value of row, by institute The M row obtainedThe two code elements direct coherent accumulation value of row is input in cellular non-coherent accumulation module, and wherein, g is for being more than In the even number of 2, M=2^d, d is integer；

(4b) g frequency domain single element coherent accumulation data stream is stored in the matrix of M row g row, g row frequency domain single element is concerned with In accumulation data stream, the numerical value of even column negates, g row frequency domain single element coherent accumulation number of the inverted for even column numerical value Carry out respective value addition according to adjacent two columns values in stream, obtain M rowTwo code elements of row process coherent accumulation value, by obtained M rowTwo code elements of row process coherent accumulation value and are input in cellular non-coherent accumulation module；

(5) cellular non-coherent accumulation is carried out:

By M rowThe two code elements direct coherent accumulation value of row and M rowTwo code elements of row process coherent accumulation value and enter respectively Row respective column data are cumulative after taking absolute value, and obtain the direct cellular non-coherent accumulation value of M row 1 row and the process cellular of M row 1 row Non-coherent accumulation value, the direct cellular non-coherent accumulation value arranged by the M row 1 obtained is incoherent with the process cellular that M row 1 arranges tired Product value is input in the peak value screening washer of peak value comparison module；

(6) peak value is compared:

(6a) the process cellular non-coherent accumulation value that direct cellular non-coherent accumulation value and the M row 1 arranged by M row 1 arranges is respectively Take the maximum in M row 1 columns value, obtain direct cellular non-coherent accumulation peak value and process cellular non-coherent accumulation peak value, will The direct cellular non-coherent accumulation peak value obtained with process in cellular non-coherent accumulation peak value input peak value comparison module In big peak comparator；

(6b) compare direct cellular non-coherent accumulation peak value and process cellular non-coherent accumulation peak value, choosing direct cellular non- The peak value that coherent accumulation peak value is maximum with process cellular non-coherent accumulation peak value, obtains the maximum of cellular non-coherent accumulation Peak value, is input to the peak-peak of the cellular non-coherent accumulation obtained in judging module；

(7) judge that the peak-peak of cellular non-coherent accumulation whether more than decision threshold, the most then performs step (8)；If No, then perform step (2)；

(8) capture is completed.

The present invention compared with prior art has the advantage that

First, due to single element accumulated value in the long code capture systems based on partial matched filter fft algorithm of the present invention Buffer module, utilizes the single element coherent accumulation that single element accumulated value buffer module is received by table tennis random access memory ram The rate reduction of data stream, overcomes baseband signal and local pseudo-code part in the technology of long code capture systems of the prior art Directly data being carried out fast Fourier transform FFT after matched filtering, the expense causing hardware resource is big, operand is big lacks Point so that the system of the present invention can save the hardware resource of fast Fourier transform FFT, reduces fast Fourier transform FFT Operand.

Second, due to the two direct phases of code element in the long code capture systems based on partial matched filter fft algorithm of the present invention Dry accumulation module, two code elements process coherent accumulation module, cellular non-coherent accumulation module, can increase according to acquisition probability and add deduct The number of few accumulation code element and cellular, overcome in prior art the parameter of device in the technology of long code capture systems cannot from Adapt to adjust, the shortcoming that capacity of resisting disturbance is low so that the system of the present invention can allow system have self-adaptative adjustment, and carries The capacity of resisting disturbance of high system.

3rd, owing to, in the long code capture systems based on partial matched filter fft algorithm of the present invention, cellular is incoherent tired The long-pending non-coherent accumulation using matrix cellular, overcomes technology Satellite navigation signal and the basis of prior art long code capture systems Ground pseudo-code linear frequency domain correlation and linear differential coherence integration, thus the long Code acquisition data of satellite navigation signals The shortcoming that processing speed is slow so that the system of the present invention improves the long Code acquisition data of satellite navigation signals and processes speed Degree.

4th, due to based on partial matched filter fft algorithm the long code capture method of the present invention, utilize multi-code relevant tired Integration method realizes the cumulative of the different weights of multiple-symbol segmentation coherent accumulation value, thus increases coherent accumulation Baud Length with relevant The time of accumulation, improve the signal to noise ratio of capture navigation signal, utilize multi cell non-coherent accumulation algorithm to realize the non-of multi cell Coherent accumulation, thus increase the number of times of non-coherent accumulation, again improve the signal to noise ratio of capture navigation signal, overcome prior art In the case of in the technology of long code capture method, Doppler frequency shift is relatively big, there is bigger being correlated with in partial matched filter fft algorithm Loss, in the case of signal to noise ratio is low, poor and circular shifting the operand of partial matched filter fft algorithm acquisition performance is big Shortcoming so that use and The inventive method achieves the long code of satellite navigation signals under maximum Doppler frequency shift, low signal-to-noise ratio environment Capture.

Accompanying drawing explanation

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

Fig. 2 is the block diagram of the signal reception memory module of the present invention；

Fig. 3 is the block diagram of the segmentation coherent accumulation module of the present invention；

Fig. 4 is the block diagram of the peak value comparison module of the present invention；

Fig. 5 is the flow chart of the inventive method；

Fig. 6 is the simulated effect figure of the present invention.

Detailed description of the invention

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

Referring to the drawings 1, the system of the present invention is further described.

The system of the present invention includes ten modules: signal receives memory module, local pseudo-code memory module, part coupling filter Mode block, single element accumulated value buffer module, fast Fourier transform FFT module, two code elements direct coherent accumulation module, two yards Unit processes coherent accumulation module, cellular non-coherent accumulation module, peak value comparison module and judging module.Wherein, signal receives and deposits Storage module, local pseudo-code memory module, partial matched filter module, fast Fourier transform FFT module, peak value comparison module and Judging module is shared module；Single element accumulated value buffer module, two code elements direct coherent accumulation module, two code elements process relevant Accumulation module and cellular non-coherent accumulation module belong to the private module of the present invention.

Signal receives memory module, for being received the multiplier in memory module by signal, signal receives storage mould The satellite navigation signals that in block, antenna receives receives, with signal, the local oscillation signal that in memory module, voltage controlled oscillator produces and is multiplied, To intermediate frequency satellite navigation signals, the intermediate frequency satellite navigation signals obtained is input to signal and receives simulation/number in memory module In word A/D sampler.With 4 times of bit rates of intermediate frequency satellite navigation signals, intermediate frequency satellite navigation signals is sampled, obtain in numeral Frequently satellite navigation signals, is input to the digital intermediate frequency satellite navigation signals obtained signal and receives bandpass filtering in memory module Device.Digital intermediate frequency satellite navigation signals is carried out baseband filtering, obtain with Doppler frequency deviation and white Gaussian noise base band Signal, is input to dividing of partial matched filter module by the baseband signal with Doppler frequency deviation and white Gaussian noise obtained In section device 1.

Local pseudo-code memory module, for being input to the local pseudo-code of the baseband signal of local pseudo-code memory module storage In the sectionaliser 2 of partial matched filter module.

Partial matched filter module, for dividing length A of the baseband signal with Doppler frequency deviation and white Gaussian noise Become B section, a length of C of subsegment of baseband signal, obtain subsegment baseband signal D, subsegment baseband signal D obtained is input to portion Dividing in the correlator in matched filtering module, wherein, the value of A is 204600, B=2^d, d is positive integer, and C isInteger Part.Use parallel search method, withFor step-length, the base band letter that in selected part matched filtering module, sectionaliser 2 receives Number the code phase of local pseudo-code, length E of the local pseudo-code of baseband signal is divided into B section, the subsegment of local pseudo-code is a length of F, obtains subsegment this locality pseudo-code G, subsegment this locality pseudo-code G obtained is input in the correlator of partial matched filter module, Wherein, the value of E is 204600, B=2^d, d is positive integer, and F isInteger part.By subsegment baseband signal D and subsegment Local pseudo-code G carries out related calculation, and obtains single element coherent accumulation data stream；By defeated for the single element coherent accumulation data stream that obtained Enter in single element accumulated value buffer module.

Single element accumulated value buffer module, is used for utilizing table tennis random access memory ram to be buffered by single element accumulated value The rate reduction of the single element coherent accumulation data stream that module receives, the single element coherent accumulation data stream after being buffered, will Single element coherent accumulation data stream after the buffering obtained is input in fast Fourier transform FFT module.

Fast Fourier transform FFT module, the single element after the buffering that fast Fourier transform FFT module is received Coherent accumulation data stream carries out L point quick Fourier conversion FFT, obtains frequency domain single element coherent accumulation data stream, will be obtained Frequency domain single element coherent accumulation data stream be input to two code elements direct coherent accumulation module neutralize two code elements process coherent accumulations In module.

Two code elements direct coherent accumulation module, for being stored in M row g row by g frequency domain single element coherent accumulation data stream Adjacent two columns values in g row frequency domain single element coherent accumulation data stream are carried out respective value addition, obtain M row by matrixRow Two code elements direct coherent accumulation value, the M row that will be obtainedIt is incoherent that the two code elements direct coherent accumulation value of row is input to cellular In accumulation module, wherein, g is the even number more than or equal to 2, M=2^d, d is integer.

Two code elements process coherent accumulation module, for g frequency domain single element coherent accumulation data stream is stored in M row g row Matrix, negates, of the inverted for even column numerical value the numerical value of even column in g row frequency domain single element coherent accumulation data stream In g row frequency domain single element coherent accumulation data stream, adjacent two columns values carry out respective value addition, obtain M rowAt two code elements of row Reason coherent accumulation value, the M row that will be obtainedTwo code elements of row process coherent accumulation value and are input to cellular non-coherent accumulation module In.

Cellular non-coherent accumulation module, for by M rowThe two code elements direct coherent accumulation value of row and M rowThe two of row Code element processes coherent accumulation value and carries out after respective column data take absolute value cumulative respectively, and the direct cellular obtaining M row 1 row is incoherent The process cellular non-coherent accumulation value of accumulated value and M row 1 row, the direct cellular non-coherent accumulation value that the M row 1 obtained is arranged with The cellular non-coherent accumulation value that processes of M row 1 row is input in the peak value screening washer of peak value comparison module.

Peak value comparison module, the non-phase of process cellular that the direct cellular non-coherent accumulation value for being arranged by M row 1 arranges with M row 1 Dry accumulated value takes the maximum in M row 1 columns value respectively, obtains direct cellular non-coherent accumulation peak value incoherent with process cellular Accumulation peak, inputs peakedness ratio relatively by the direct cellular non-coherent accumulation peak value obtained with processing cellular non-coherent accumulation peak value In peak-peak comparator in module.By direct cellular non-coherent accumulation peak value and process cellular non-coherent accumulation peakedness ratio Relatively, choose direct cellular non-coherent accumulation peak value and process a peak value maximum in cellular non-coherent accumulation peak value, obtaining unit The peak-peak of born of the same parents' non-coherent accumulation, is input to the peak-peak of the cellular non-coherent accumulation obtained in judging module.

Judging module, for the peak-peak of cellular non-coherent accumulation is compared with decision threshold, if no more than decision gate Limit, then the code phase changing the local pseudo-code of search re-searches for；If more than decision threshold, then complete capture.

Referring to the drawings 2, the signal of the present invention receives memory module and includes 5 devices: antenna, voltage controlled oscillator, multiplier, Analog/digital A/D sampler and band filter；Wherein, antenna, voltage controlled oscillator, multiplier, analog/digital A/D sampler It is public device with band filter.

Antenna is for receiving satellite navigation signals.

Voltage controlled oscillator is the local oscillation signal carrying out mixing operations for generation and navigation signal.

Multiplier is for signal receives the satellite navigation signals and signal reception storage that in memory module, antenna receives The local oscillation signal that in module, voltage controlled oscillator produces is multiplied, and obtains intermediate frequency satellite navigation signals, is led by the intermediate frequency satellite obtained Boat signal is input to signal and receives in memory module in analog/digital A/D sampler.

Analog/digital A/D sampler is to believe intermediate frequency satellite navigation for 4 times of bit rates with intermediate frequency satellite navigation signals Number sampling, obtain digital intermediate frequency satellite navigation signals, the digital intermediate frequency satellite navigation signals obtained is input to signal receive Band filter in memory module.

Band filter is for digital intermediate frequency satellite navigation signals being carried out baseband filtering, obtaining with Doppler frequency deviation With white Gaussian noise baseband signal, the baseband signal with Doppler frequency deviation and white Gaussian noise obtained is input to In the sectionaliser 1 of partial matched filter module.

Referring to the drawings 3, the partial matched filter module of the present invention includes 3 devices: sectionaliser 1, sectionaliser 2 and relevant Device；Wherein, sectionaliser 1, sectionaliser 2 and correlator are public devices.

Sectionaliser 1 is for length A of the baseband signal with Doppler frequency deviation and white Gaussian noise is divided into B section, base The a length of C of subsegment of band signal, obtains subsegment baseband signal D, and subsegment baseband signal D obtained is input to part coupling filter In correlator in mode block, wherein, the value of A is 204600, B=2^d, d is positive integer, and C isInteger part.

Sectionaliser 2 is for using parallel search method, withFor step-length, segmentation in selected part matched filtering module The code phase of the local pseudo-code of the baseband signal that device 2 receives, is divided into B section by length E of the local pseudo-code of baseband signal, local pseudo- The a length of F of subsegment of code, obtains subsegment this locality pseudo-code G, subsegment this locality pseudo-code G obtained is input to partial matched filter mould In the correlator of block, wherein, the value of E is 204600, B=2^d, d is positive integer, and F isInteger part.

Correlator is subsegment baseband signal D to be carried out related calculation with subsegment this locality pseudo-code G, obtains single element coherent accumulation number According to stream, the single element coherent accumulation data stream obtained is input in single element accumulated value buffer module.

Referring to the drawings 4, the peak value comparison module of the present invention includes 2 devices: peak value screening washer and peak-peak comparator； Wherein, peak value screening washer and peak-peak comparator are public devices.

Peak value screening washer is the non-phase of process cellular arranged with M row 1 for the direct cellular non-coherent accumulation value arranged by M row 1 Dry accumulated value takes the maximum in M row 1 columns value respectively, obtains direct cellular non-coherent accumulation peak value incoherent with process cellular Accumulation peak, inputs peakedness ratio relatively mould by the direct cellular non-coherent accumulation peak value obtained with processing cellular non-coherent accumulation peak value In peak-peak comparator in block, wherein, M=2^d, d is integer.

Peak-peak comparator is for by direct cellular non-coherent accumulation peak value and process cellular non-coherent accumulation peak value Relatively, choose direct cellular non-coherent accumulation peak value and process a peak value maximum in cellular non-coherent accumulation peak value, obtaining The peak-peak of cellular non-coherent accumulation, is input to the peak-peak of the cellular non-coherent accumulation obtained in judging module.

Below in conjunction with the accompanying drawings 5, the method for the present invention is described further.

Step 1, it is thus achieved that baseband signal.

Receive the multiplier in memory module by signal, signal is received the satellite navigation that in memory module, antenna receives Signal h (t) receives, with signal, local oscillation signal i (t) that in memory module, voltage controlled oscillator produces and is multiplied, and obtains intermediate frequency satellite navigation Signal j (t), is input to intermediate frequency satellite navigation signals j (t) obtained analog/digital A/D in signal reception memory module and adopts In sample device, wherein t is the moment of satellite navigation signals.

Intermediate frequency satellite navigation signals j (t) is sampled by the present invention with 4 times of bit rates of intermediate frequency satellite navigation signals j (t), To digital intermediate frequency satellite navigation signals u (n), digital intermediate frequency satellite navigation signals u (n) obtained is input to signal reception and deposits Band filter in storage module.

Digital intermediate frequency satellite navigation signals u (n) is carried out baseband filtering, obtains with Doppler frequency deviation and white Gaussian noise Baseband signal r (n), baseband signal r (n) with Doppler frequency deviation and white Gaussian noise obtained is input to part In the sectionaliser 1 of matched filtering module, baseband signal r (n) is expressed as:

R (n)=Ad (n) p (n+ τ) e^j2πfnΔT+N

Wherein, r (n) represents that n represents the sampling instant of baseband signal, and A represents in the baseband signal that sampling instant n receives The amplitude of baseband signal, d (n) represents the modulated symbol in sampling instant n baseband signal, and p (n+ τ) represents at sampling instant n base The long pseudo-random code of band signal, τ represents the offset phase of the long pseudo-random code of baseband signal, and e represents that natural constant, j represent empty Number unit, π represents that pi, f represent that baseband signal Doppler frequency deviation, Δ T represent the sampling time interval of baseband signal, N table Show the white Gaussian noise of baseband signal.

Step 2, carries out partial matched filter computing.

Length A of baseband signal r (n) with Doppler frequency deviation and white Gaussian noise is divided into B section, the son of baseband signal Segment length is C, obtains subsegment baseband signal D, subsegment baseband signal D obtained is input in partial matched filter module In correlator, wherein, the value of A is 204600, B=2^d, d is positive integer, and C isInteger part.

Local pseudo-code q (n) of the baseband signal of local pseudo-code memory module storage is input to partial matched filter module In middle sectionaliser 2.

Use parallel search method, withFor step-length, the base band that in selected part matched filtering module, sectionaliser 2 receives The code phase of the local pseudo-code of signal, wherein, it is the method using parallel search that local pseudo-code takes any code phase α, and this is parallel The method of search chooses the phase place parallel computation of the local pseudo-code of 8 baseband signals, simultaneously 8 phase places of detection simultaneously, by base band Length E of local pseudo-code q (n+ α) of signal is divided into B section, a length of F of subsegment of local pseudo-code, obtains subsegment this locality pseudo-code G, will Subsegment this locality pseudo-code G obtained is input in the correlator of partial matched filter module, and wherein, the value of E is 204600, B=2^d, d is positive integer, and F isInteger part.

Utilize following formula, subsegment baseband signal D is carried out related calculation with subsegment this locality pseudo-code G, obtains single element coherent accumulation Data stream s (m).

$s\left(m\right)=\underset{n=\mathrm{mC}}{\overset{(m+1)C-1}{\mathrm{\Σ}}}r\left(n\right)q(n+\mathrm{\α})$

Wherein, single element coherent accumulation data when s (m) represents section serial number m, m represents single element coherent accumulation data Section sequence number, wherein m=0,1 ..., B-1, B represent the single element segments of baseband signal, and ∑ represents that sum operation, C represent base The subsegment length of band signal, r (n) represents that n represents the sampling instant of baseband signal, q in the baseband signal that sampling instant n receives (n+ α) represents the local pseudo-code in sampling instant n baseband signal, and α represents the local pseudo-code initial ranging code phase of baseband signal.

Single element coherent accumulation data stream s (m) obtained is input in single element accumulated value buffer module.

Step 3, carries out fast Fourier transform FFT.

Utilize the single element coherent accumulation number that single element accumulated value buffer module is received by table tennis random access memory ram According to the rate reduction of stream s (m), the single element coherent accumulation data stream s after being buffered_h(m), the list after the buffering that will be obtained Code element coherent accumulation data stream s_hM () is input in fast Fourier transform FFT module.

Utilize following formula, the single element coherent accumulation data stream s after the buffering receive fast Fourier transform FFT module_h M () carries out L point quick Fourier conversion FFT, obtain frequency domain single element coherent accumulation data stream S.

s₁(β)=s_h(2β)

s₂(β)=s_h(2β+1)

$S_1\left(k\right)=\underset{\mathrm{\β}=0}{\overset{\frac{\mathrm{\β}}{2}-1}{\mathrm{\Σ}}}{s}_1\left(\mathrm{\β}\right)e^{\frac{-j4\mathrm{\πk\β}}{B}}$

$S_2\left(k\right)=\underset{\mathrm{\β}=0}{\overset{\frac{B}{2}-1}{\mathrm{\Σ}}}{s}_2\left(\mathrm{\β}\right)e^{\frac{-j4\mathrm{\πk\β}}{B}}$

$S\left(k\right)=S_1\left(k\right)+e^{\frac{-j2\mathrm{\πk}}{L}}{S}_2\left(k\right)$

$S(k+\frac{L}{2})=S_1\left(k\right)-e^{\frac{-j2\mathrm{\πk}}{L}}{S}_2\left(k\right)$

Wherein, s₁(β) the single element coherent accumulation data after even number section sequence number buffering when first half section serial number β are represented Stream, β represents the first half section sequence number of single element coherent accumulation data, whereinB represents the solid size of baseband signal Unit's segments, s_h(2 β) represents the single element coherent accumulation data when even number section serial number 2 β after buffering, s₂(β) represent front Single element coherent accumulation data stream after odd number section sequence number buffers during half section of serial number β, s_h(2 β+1) represents in odd number section sequence number Single element coherent accumulation data after buffering when being 2 β+1, ∑ represents sum operation, and S (k) represents at half frequency domain sample point k BeforeThe frequency spectrum of item single element coherent accumulation data stream, k represents the half frequency domain sample point of single element coherent accumulation data stream, Wherein,S₁K () represents the even number section sequence number single element coherent accumulation data at half frequency domain sample point k The frequency spectrum of stream, e represents that natural constant, j represent that imaginary unit, L represent counting of fast Fourier transform FFT, wherein L=2^d, d For positive integer, S₂K () represents the frequency spectrum of the odd number section sequence number single element coherent accumulation data stream at half frequency domain sample point k,Represent at half frequency domain sample pointAfter ChuThe frequency spectrum of item single element coherent accumulation data stream.

The frequency domain single element coherent accumulation data stream S obtained is input to two code elements direct coherent accumulation module and neutralizes two Code element processes in coherent accumulation module.

Step 4, carries out the coherent accumulation of two code element different weights.

G frequency domain single element coherent accumulation data stream S is stored in the matrix of M row g row, relevant tired to g row frequency domain single element In volume data stream, adjacent two columns values carry out respective value addition, obtain M rowTwo code elements direct coherent accumulation value Y of row, by institute The M row obtainedTwo code elements direct coherent accumulation value Y of row is input in cellular non-coherent accumulation module.

G frequency domain single element coherent accumulation data stream S is stored in the matrix of M row g row, relevant tired to g row frequency domain single element In volume data stream S, the numerical value of even column negates, g row frequency domain single element coherent accumulation data of the inverted for even column numerical value In stream, adjacent two columns values carry out respective value addition, obtain M rowTwo code elements of row process coherent accumulation value Z, the M that will be obtained OKTwo code elements of row process coherent accumulation value Z and are input in cellular non-coherent accumulation module.

Step 5, carries out cellular non-coherent accumulation.

By M rowTwo code elements direct coherent accumulation value Y of row and M rowTwo code elements of row process coherent accumulation value Z respectively Carry out after respective column data take absolute value cumulative, obtain direct cellular non-coherent accumulation value V of M row 1 row and the process of M row 1 row Cellular non-coherent accumulation value W, direct cellular non-coherent accumulation value V arranged by the M row 1 obtained is non-with the process cellular that M row 1 arranges Coherent accumulation value W is input in the peak value screening washer of peak value comparison module, and wherein, cellular non-coherent accumulation is to use matrix cellular Non-coherent accumulation, save hardware resource cost, improve capture speed.

Step 6, compares peak value.

Process cellular non-coherent accumulation value W that direct cellular non-coherent accumulation value V and the M row 1 arranged by M row 1 arranges takes M respectively Maximum in row 1 columns value, obtains direct cellular non-coherent accumulation peak value V_maxWith process cellular non-coherent accumulation peak value W_max, The direct cellular non-coherent accumulation peak value V that will be obtained_maxWith process cellular non-coherent accumulation peak value W_maxInput peakedness ratio relatively mould In peak-peak comparator in block.

By direct cellular non-coherent accumulation peak value V_maxWith process cellular non-coherent accumulation peak value W_maxRelatively, direct element is chosen Born of the same parents non-coherent accumulation peak value V_maxWith process cellular non-coherent accumulation peak value W_maxOne peak value of middle maximum, obtains cellular incoherent Peak-peak X of accumulation, is input to peak-peak X of the cellular non-coherent accumulation obtained in judging module.

Step 7, it is judged that the peak-peak of cellular non-coherent accumulation whether more than decision threshold, the most then performs step 8； If it is not, then perform step 2.

Step 8, completes capture.

Below in conjunction with the accompanying drawings 6, the simulated effect of the present invention is described further.

1. simulated conditions:

The emulation of the present invention uses Matlab7.10 simulation software, and the bit rate of modulation code is 50 bits per second, pseudorandom The bit rate of long code is 10.23 megabits per second, and the modulation system of signal is binary phase shift keying BPSK.Satellite navigation signals SNR ranges be-55dB to-30dB, the long code acquisition probability decision threshold of satellite navigation signals is set to 0.7, satellite navigation The Doppler frequency deviation of signal is [0Hz, 5Hz, 50Hz, 500Hz, 1000Hz, 5000Hz], the long Code acquisition side of satellite navigation signals In method, the code element number of multiple-symbol coherent accumulation is 2, and the cellular number of multi cell non-coherent accumulation is 2, and the single element of baseband signal divides Hop count B=256, subsegment length C=800 of baseband signal, counting of fast Fourier transform FFT is 256.

2. emulation content:

Fig. 6 (a) uses the long code capture method of the inventive method and partial matched filter FFT satellite navigation signals to exist respectively Doppler frequency deviation is 0Hz, the SNR ranges of satellite navigation signals is the length obtaining satellite navigation signals during-55dB to-30dB Code acquisition probability, and compare both approaches long code acquisition performance of satellite navigation signals under low signal-to-noise ratio environment.

Fig. 6 (b) use the inventive method respectively Doppler frequency deviation be [0Hz, 5Hz, 50Hz, 500Hz, 1000Hz, 5000Hz], the SNR ranges of satellite navigation signals be that to obtain the long Code acquisition of satellite navigation signals during-55dB to-30dB general Rate.

The long code capture method of Fig. 6 (c) employing partial matched filter FFT satellite navigation signals at Doppler frequency deviation is respectively Ask when the SNR ranges of [0Hz, 5Hz, 50Hz, 500Hz, 1000Hz, 5000Hz], satellite navigation signals is for-55dB to-30dB Go out the long code acquisition probability of satellite navigation signals.

3. analysis of simulation result:

Fig. 6 (a) is that how general the nothing of the long code capture method of the inventive method and partial matched filter FFT satellite navigation signals is Strangle frequency deviation long code acquisition performance comparison diagram.In Fig. 6 (a) withThe curve that line indicates represents that the nothing using the inventive method is many General Le frequency deviation long code acquisition probability.In Fig. 6 (a) withThe curve that line indicates represents that employing partial matched filter FFT satellite is led The acquisition probability of long code without Doppler frequency deviation of the long code capture method of boat signal.

From Fig. 6 (a) it can be seen that withLine indicate and withThe present invention on two curve overall trends that line indicates Method acquisition probability when the signal to noise ratio of satellite navigation signals is for-40dB just reaches 1, and partial matched filter FFT satellite navigation The long code capture method of signal acquisition probability when the signal to noise ratio of satellite navigation signals is for-35dB just reaches 1, it is seen that side of the present invention Method has broader low signal-to-noise ratio subject range than the long code capture method of partial matched filter FFT satellite navigation signals, in phase Under the conditions of same long code acquisition probability, the inventive method is than the long code capture method of partial matched filter FFT satellite navigation signals Performance improves about 5dB, and the inventive method is suitable for the long code of satellite navigation signals under the low signal-to-noise ratio environment of more than-40dB and catches Obtain.

Fig. 6 (b) is that the Doppler frequency deviation long code that has of the long code capture method of partial matched filter FFT satellite navigation signals is caught Obtain performance chart.In Fig. 6 (b) withThe curve that line indicates represents the length of partial matched filter FFT satellite navigation signals Code capture method is at the long code acquisition probability that Doppler frequency deviation is during 0Hz.In Fig. 6 (b) withThe curve expressed portion that line indicates The long code capture method dividing matched filtering FFT satellite navigation signals is long code acquisition probability during 5Hz at Doppler frequency deviation.Fig. 6 In (b) withThe curve that line indicates represents that the long code capture method of partial matched filter FFT satellite navigation signals is Doppler Frequency deviation is long code acquisition probability during 50Hz.In Fig. 6 (b) withThe curve that line indicates represents partial matched filter FFT satellite The long code capture method of navigation signal is at the long code acquisition probability that Doppler frequency deviation is during 500Hz.In Fig. 6 (b) withLine mark The curve shown represents that the long code capture method of partial matched filter FFT satellite navigation signals is when Doppler frequency deviation is 1000Hz Long code acquisition probability.In Fig. 6 (b) withThe curve that line indicates represents the length of partial matched filter FFT satellite navigation signals Code capture method is at the long code acquisition probability that Doppler frequency deviation is during 5000Hz.

From Fig. 6 (b) it can be seen that withLine indicate, withLine indicate, withLine indicate, withLine mark Show, withLine indicate and withThe six roots of sensation curve overall trend upper part matched filtering FFT satellite navigation letter that line indicates Number long code capture method when Doppler frequency deviation is 0Hz, 5Hz, 50Hz, 500Hz, 1000Hz Doppler frequency deviation to satellite navigation The long code acquisition probability impact of signal is little.The signal to noise ratio of satellite navigation signals is about when-35dB, and the method can capture completely Long code to satellite navigation signals.But the long code capture method of partial matched filter FFT satellite navigation signals is at Doppler frequency deviation The long code acquisition probability of satellite navigation signals is affected very big for Doppler frequency deviation during 5000Hz, the signal to noise ratio of satellite navigation signals When-30dB, the long code acquisition probability of the satellite navigation signals of the method is 0.88.Visible in large doppler frequency deviation, low noise Satellite navigation letter high performance can not be captured than the long code capture method of environment lower part matched filtering FFT satellite navigation signals Number long code.

Fig. 6 (c) be the inventive method have Doppler frequency deviation long code acquisition performance curve chart.In Fig. 6 (c) withLine The curve indicated represents that the inventive method is at the long code acquisition probability that Doppler frequency deviation is during 0Hz.In Fig. 6 (c) withLine mark The curve shown represents that the inventive method is at the long code acquisition probability that Doppler frequency deviation is during 5Hz.In Fig. 6 (c) withLine indicates Curve represent that the inventive method is at the long code acquisition probability that Doppler frequency deviation is during 50Hz.In Fig. 6 (c) withLine indicates Curve represent that the inventive method is at the long code acquisition probability that Doppler frequency deviation is during 500Hz.In Fig. 6 (c) withLine indicates Curve represent that the inventive method is at the long code acquisition probability that Doppler frequency deviation is during 1000Hz.In Fig. 6 (c) withLine mark The curve shown represents that the inventive method is at the long code acquisition probability that Doppler frequency deviation is during 5000Hz.

From Fig. 6 (c) it can be seen that withLine indicate, withLine indicate, withLine indicate, withLine mark Show, withLine indicate and withOn the six roots of sensation curve overall trend that line indicates, the inventive method at Doppler frequency deviation is When 0Hz, 5Hz, 50Hz, 500Hz, 1000Hz, the long code acquisition probability of satellite navigation signals is affected little by Doppler frequency deviation, defends The signal to noise ratio of star navigation signal is about when-40dB, and the inventive method can capture the long code of satellite navigation signals completely；Simultaneously The long code acquisition probability of satellite navigation signals is affected also by the inventive method Doppler frequency deviation when Doppler frequency deviation is 5000Hz Very big, but the signal to noise ratio of satellite navigation signals is when-30dB, and the long code acquisition probability of the satellite navigation signals of the inventive method is 0.98；Visible under large doppler frequency deviation, low signal-to-noise ratio environment the inventive method than partial matched filter FFT satellite navigation signals The performance of long code of capture satellite navigation signals of long code capture method high.

Claims (5)

1. a long code capture systems based on partial matched filter fft algorithm, receives memory module, local pseudo-code including signal Memory module, partial matched filter module, single element accumulated value buffer module, fast Fourier transform FFT module, two code elements are straight Connect coherent accumulation module, two code elements process coherent accumulation modules, cellular non-coherent accumulation module, peak value comparison module and judgement mould Block；Wherein:

Described signal receives memory module, for being received the multiplier in memory module by signal, signal is received storage The satellite navigation signals that in module, antenna receives receives, with signal, the local oscillation signal that in memory module, voltage controlled oscillator produces and is multiplied, Obtain intermediate frequency satellite navigation signals, the intermediate frequency satellite navigation signals obtained is input to signal receive in memory module simulate/ In numeral A/D sampler；With 4 times of bit rates of intermediate frequency satellite navigation signals, intermediate frequency satellite navigation signals is sampled, obtain numeral Intermediate frequency satellite navigation signals, is input to the digital intermediate frequency satellite navigation signals obtained signal and receives the logical filter of band in memory module Ripple device；Digital intermediate frequency satellite navigation signals is carried out baseband filtering, obtains the base band with Doppler frequency deviation and white Gaussian noise Signal, is input to dividing of partial matched filter module by the baseband signal with Doppler frequency deviation and white Gaussian noise obtained In section device 1；

Described local pseudo-code memory module, for the local pseudo-code input by the baseband signal of local pseudo-code memory module storage In the sectionaliser 2 of partial matched filter module；

Described partial matched filter module, for by length A of the baseband signal with Doppler frequency deviation and white Gaussian noise It is divided into B section, a length of C of subsegment of baseband signal, obtain subsegment baseband signal D, subsegment baseband signal D obtained is input to In correlator in partial matched filter module, wherein, the value of A is 204600, B=2^d, d is positive integer, and C isWhole Fractional part；Use parallel search method, withFor step-length, the base band that in selected part matched filtering module, sectionaliser 2 receives The code phase of the local pseudo-code of signal, is divided into B section, the subsegment length of local pseudo-code by length E of the local pseudo-code of baseband signal For F, obtain subsegment this locality pseudo-code G, subsegment this locality pseudo-code G obtained is input to the correlator of partial matched filter module In, wherein, the value of E is 204600, B=2^d, d is positive integer, and F isInteger part；By subsegment baseband signal D and son Local pseudo-code G of section carries out related calculation, and obtains single element coherent accumulation data stream；The single element coherent accumulation data stream that will be obtained It is input in single element accumulated value buffer module；

Described single element accumulated value buffer module, is used for utilizing table tennis random access memory ram to be delayed by single element accumulated value The rate reduction of the single element coherent accumulation data stream that die block receives, the single element coherent accumulation data stream after being buffered, Single element coherent accumulation data stream after the buffering that will be obtained is input in fast Fourier transform FFT module；

Described fast Fourier transform FFT module, the solid size after the buffering that fast Fourier transform FFT module is received Unit's coherent accumulation data stream carries out L point quick Fourier conversion FFT, obtains frequency domain single element coherent accumulation data stream；To be obtained Frequency domain single element coherent accumulation data stream be input to two code elements direct coherent accumulation module and neutralize two code elements and process relevant tired In volume module；

Described two code elements direct coherent accumulation module, for being stored in M row g row by g frequency domain single element coherent accumulation data stream Matrix, adjacent two columns values in g row frequency domain single element coherent accumulation data stream are carried out respective value addition, obtain M rowRow Two code elements direct coherent accumulation value, the M row that will be obtainedThe two code elements direct coherent accumulation value of row is input to the non-phase of cellular In dry accumulation module, wherein, g is the even number more than or equal to 2, M=2^d, d is integer；

Two described code elements process coherent accumulation module, for g frequency domain single element coherent accumulation data stream is stored in M row g row Matrix, the numerical value of even column in g row frequency domain single element coherent accumulation data stream is negated, after even column numerical value is negated G row frequency domain single element coherent accumulation data stream in adjacent two columns values carry out respective value addition, obtain M rowTwo code elements of row Process coherent accumulation value, the M row that will be obtainedTwo code elements of row process coherent accumulation value and are input to cellular non-coherent accumulation mould In block；

Described cellular non-coherent accumulation module, for by M rowThe two code elements direct coherent accumulation value of row and M rowThe two of row Code element processes coherent accumulation value and carries out after respective column data take absolute value cumulative respectively, and the direct cellular obtaining M row 1 row is incoherent The process cellular non-coherent accumulation value of accumulated value and M row 1 row, the direct cellular non-coherent accumulation value that the M row 1 obtained is arranged with The cellular non-coherent accumulation value that processes of M row 1 row is input in the peak value screening washer of peak value comparison module；

Described peak value comparison module, non-with the process cellular that M row 1 arranges for the direct cellular non-coherent accumulation value that M row 1 is arranged Coherent accumulation value takes the maximum in M row 1 columns value respectively, obtains direct cellular non-coherent accumulation peak value and processes the non-phase of cellular Dry accumulation peak, inputs peakedness ratio by the direct cellular non-coherent accumulation peak value obtained with processing cellular non-coherent accumulation peak value In peak-peak comparator in relatively module；By direct cellular non-coherent accumulation peak value and process cellular non-coherent accumulation peakedness ratio Relatively, choose direct cellular non-coherent accumulation peak value and process a peak value maximum in cellular non-coherent accumulation peak value, obtaining unit The peak-peak of born of the same parents' non-coherent accumulation, is input to the peak-peak of the cellular non-coherent accumulation obtained in judging module；

Described judging module, for comparing the peak-peak of cellular non-coherent accumulation with decision threshold, if being not more than judgement Thresholding, then the code phase changing the local pseudo-code of search re-searches for；If more than decision threshold, then complete capture.

Long code capture systems based on partial matched filter fft algorithm the most according to claim 1, it is characterised in that institute The partial matched filter module stated includes sectionaliser 1, sectionaliser 2 and correlator；Wherein,

Described sectionaliser 1 is for being divided into B section, base by length A of the baseband signal with Doppler frequency deviation and white Gaussian noise The a length of C of subsegment of band signal, obtains subsegment baseband signal D, and subsegment baseband signal D obtained is input to part coupling filter In correlator in mode block, wherein, the value of A is 204600, B=2^d, d is positive integer, and C isInteger part；

Described sectionaliser 2 is used for using parallel search method, withFor step-length, segmentation in selected part matched filtering module The code phase of the local pseudo-code of the baseband signal that device 2 receives, is divided into B section by length E of the local pseudo-code of baseband signal, local pseudo- The a length of F of subsegment of code, obtains subsegment this locality pseudo-code G, subsegment this locality pseudo-code G obtained is input to partial matched filter mould In the correlator of block, wherein, the value of E is 204600, B=2^d, d is positive integer, and F isInteger part；

Subsegment baseband signal D is carried out related calculation by described correlator with subsegment this locality pseudo-code G, obtains single element coherent accumulation number According to stream, the single element coherent accumulation data stream obtained is input in single element accumulated value buffer module.

Long code capture systems based on partial matched filter fft algorithm the most according to claim 1, it is characterised in that institute The peak value comparison module stated includes peak value screening washer and peak-peak comparator；Wherein,

The non-phase of process cellular that described peak value screening washer arranges with M row 1 for the direct cellular non-coherent accumulation value arranged by M row 1 Dry accumulated value takes the maximum in M row 1 columns value respectively, obtains direct cellular non-coherent accumulation peak value incoherent with process cellular Accumulation peak, inputs peakedness ratio relatively mould by the direct cellular non-coherent accumulation peak value obtained with processing cellular non-coherent accumulation peak value In peak-peak comparator in block, wherein, M=2^d, d is integer；

Described peak-peak comparator is for by direct cellular non-coherent accumulation peak value and process cellular non-coherent accumulation peak value Relatively, choose direct cellular non-coherent accumulation peak value and process a peak value maximum in cellular non-coherent accumulation peak value, obtaining The peak-peak of cellular non-coherent accumulation, is input to the peak-peak of the cellular non-coherent accumulation obtained in judging module.

4. a long code capture method based on partial matched filter fft algorithm, comprises the following steps that

(1) baseband signal is obtained:

(1a) receive the multiplier in memory module by signal, signal is received the satellite navigation that in memory module, antenna receives Signal receives, with signal, the local oscillation signal that in memory module, voltage controlled oscillator produces and is multiplied, and obtains intermediate frequency satellite navigation signals, will The intermediate frequency satellite navigation signals obtained is input to signal and receives in memory module in analog/digital A/D sampler；

(1b) with 4 times of bit rates of intermediate frequency satellite navigation signals, intermediate frequency satellite navigation signals is sampled, obtain digital intermediate frequency satellite Navigation signal, is input to the digital intermediate frequency satellite navigation signals obtained signal and receives band filter in memory module；

(1c) digital intermediate frequency satellite navigation signals is carried out baseband filtering, obtain the base with Doppler frequency deviation and white Gaussian noise Band signal, is input to partial matched filter module by the baseband signal with Doppler frequency deviation and white Gaussian noise obtained In sectionaliser 1, baseband signal is expressed as:

R (n)=Ad (n) p (n+ τ) e^j2πfnΔT+N

Wherein, r (n) represents that n represents the sampling instant of baseband signal, and A represents base band in the baseband signal that sampling instant n receives The amplitude of signal, d (n) represents the modulated symbol in sampling instant n baseband signal, and p (n+ τ) represents at sampling instant n base band letter Number long pseudo-random code, τ represents the offset phase of the long pseudo-random code of baseband signal, and e represents that natural constant, j represent imaginary number list Position, π represents that pi, f represent that baseband signal Doppler frequency deviation, Δ T represent the sampling time interval of baseband signal, and N represents base The white Gaussian noise of band signal；

(2) partial matched filter computing is carried out:

(2a) length A of the baseband signal with Doppler frequency deviation and white Gaussian noise being divided into B section, the subsegment of baseband signal is long Degree is C, obtains subsegment baseband signal D, and subsegment baseband signal D obtained be input in partial matched filter module is relevant In device, wherein, the value of A is 204600, B=2^d, d is positive integer, and C isInteger part；

(2b) it is input in partial matched filter module divide by the local pseudo-code of the baseband signal of local pseudo-code memory module storage In section device 2；

(2c) parallel search method is used, withFor step-length, the base band that in selected part matched filtering module, sectionaliser 2 receives The code phase of the local pseudo-code of signal, is divided into B section, the subsegment length of local pseudo-code by length E of the local pseudo-code of baseband signal For F, obtain subsegment this locality pseudo-code G, subsegment this locality pseudo-code G obtained is input to the correlator of partial matched filter module In, wherein, the value of E is 204600, B=2^d, d is positive integer, and F isInteger part；

(2d) utilize following formula, subsegment baseband signal D is carried out related calculation with subsegment this locality pseudo-code G, obtains single element coherent accumulation Data stream；

$s\left(m\right)=\underset{n=mC}{\overset{(m+1)C-1}{\Σ}}r\left(n\right)q(n+\mathrm{\α})$

Wherein, single element coherent accumulation data when s (m) represents section serial number m, m represents the Duan Xu of single element coherent accumulation data Number, wherein m=0,1 ..., B-1, B represent the single element segments of baseband signal, and ∑ represents that sum operation, C represent that base band is believed Number subsegment length, r (n) represent sampling instant n receive baseband signal, n represents the sampling instant of baseband signal, q (n+ α) Representing the local pseudo-code in sampling instant n baseband signal, α represents the local pseudo-code initial ranging code phase of baseband signal；

(2e) the single element coherent accumulation data stream obtained is input in single element accumulated value buffer module；

(3) fast Fourier transform FFT is carried out:

(3a) the single element coherent accumulation number that single element accumulated value buffer module is received by table tennis random access memory ram is utilized According to the rate reduction of stream, the single element coherent accumulation data stream after being buffered, the single element after the buffering that will be obtained is concerned with Accumulation data stream is input in fast Fourier transform FFT module；

(3b) utilizing following formula, the single element coherent accumulation data after the buffering receive fast Fourier transform FFT module flow to Row L point quick Fourier conversion FFT, obtains frequency domain single element coherent accumulation data stream；

$S\left(k\right)=S_1\left(k\right)+e^{\frac{-j2\mathrm{\π}k}{L}}{S}_2\left(k\right)$

$S(k+\frac{L}{2})=S_1\left(k\right)-e^{\frac{-j2\mathrm{\π}k}{L}}{S}_2\left(k\right)$

Wherein, before S (k) represents at half frequency domain sample point kThe frequency spectrum of item single element coherent accumulation data stream, k represents single The half frequency domain sample point of code element coherent accumulation data stream, wherein,S₁K () represents at half frequency domain sample The frequency spectrum of the even number section sequence number single element coherent accumulation data stream at some k, e represents that natural constant, j represent that imaginary unit, L represent Fast Fourier transform FFT counts, wherein L=2^d, d is positive integer, S₂K () represents the odd number at half frequency domain sample point k The frequency spectrum of section sequence number single element coherent accumulation data stream,Represent at half frequency domain sample pointAfter ChuXiang Dan The frequency spectrum of code element coherent accumulation data stream；

(3c) the frequency domain single element coherent accumulation data stream obtained is input to two code elements direct coherent accumulation module and neutralizes two Code element processes in coherent accumulation module；

(4) coherent accumulation of two code element different weights is carried out:

(4a) g frequency domain single element coherent accumulation data stream is stored in the matrix of M row g row, to g row frequency domain single element coherent accumulation In data stream, adjacent two columns values carry out respective value addition, obtain M rowThe two code elements direct coherent accumulation value of row, will be obtained M rowThe two code elements direct coherent accumulation value of row is input in cellular non-coherent accumulation module, and wherein, g is more than or equal to 2 Even number, M=2^d, d is integer；

(4b) g frequency domain single element coherent accumulation data stream is stored in the matrix of M row g row, to g row frequency domain single element coherent accumulation In data stream, the numerical value of even column negates, g row frequency domain single element coherent accumulation data stream of the inverted for even column numerical value In adjacent two columns values carry out respective value addition, obtain M rowTwo code elements of row process coherent accumulation value, the M row that will be obtainedTwo code elements of row process coherent accumulation value and are input in cellular non-coherent accumulation module；

(5) cellular non-coherent accumulation is carried out:

By M rowThe two code elements direct coherent accumulation value of row and M rowTwo code elements of row process coherent accumulation value and carry out correspondence respectively Column data is cumulative after taking absolute value, and the direct cellular non-coherent accumulation value obtaining M row 1 row is incoherent with the process cellular that M row 1 arranges Accumulated value, the process cellular non-coherent accumulation value that direct cellular non-coherent accumulation value and the M row 1 arranged by the M row 1 obtained arranges is defeated Enter in the peak value screening washer of peak value comparison module；

(6) peak value is compared:

(6a) the process cellular non-coherent accumulation value that direct cellular non-coherent accumulation value and the M row 1 arranged by M row 1 arranges takes M row respectively Maximum in 1 columns value, obtains direct cellular non-coherent accumulation peak value and processes cellular non-coherent accumulation peak value, will be obtained Direct cellular non-coherent accumulation peak value with process cellular non-coherent accumulation peak value input peak value comparison module in peak-peak In comparator；

(6b) compare direct cellular non-coherent accumulation peak value and process cellular non-coherent accumulation peak value, choosing direct cellular incoherent The peak value that accumulation peak is maximum with process cellular non-coherent accumulation peak value, using this peak value as cellular non-coherent accumulation Peak-peak, is input to the peak-peak of the cellular non-coherent accumulation obtained in judging module；

(7) judge that the peak-peak of cellular non-coherent accumulation, whether more than decision threshold, if so, performs step (8)；Otherwise, hold Row step (2)；

(8) capture is completed.

Long code capture method based on partial matched filter fft algorithm the most according to claim 4, it is characterised in that step Suddenly the parallel search method described in (2c) refers to, chooses the phase place parallel computation of the local pseudo-code of 8 baseband signals simultaneously, with Time 8 baseband signals of detection the phase place of local pseudo-code.