CN101072045A - Short-period spreading code time frequency parallel searching method based on frequency folding - Google Patents

Abstract

Characters of the method are that carrying out folding operation for received signals in different frequencies to implement parallel searching frequency units; next, using FFT technique for implementing correlation between received signals and local pseudo code, and implementing parallel searching code phase units; when implementing, the method determines number of frequency unit to be searched based on frequency range and step length for searching frequency, and then carries out folding operation for signals on all frequency cells; afterwards, using FFT technique for implementing parallel correlation between the folded signals and local pseudo code so as to realize parallel searching multiple code phases. Under condition of not increasing complexity of algorithm and hardware markedly, realizing parallel searching carrier frequency unit and multiple code phase units, the method raises searching speed, and reduces time for capturing pseudo code.

Description

Short-period spreading code time frequency parallel searching method based on frequency folding

Technical field

The invention belongs to spread spectrum communication, relate in particular to the spread spectrum code acquisition technical field.

Background technology

Spread spectrum communication has lot of advantages, uses very wide.The adverse effect that spread-spectrum signal can resist or suppress to disturb, for example interference that other users cause in the channel; Spread-spectrum signal sends hidden signal with low-power, makes that signal is difficult to be detected by the listener-in; Can realize secure communication.Spread spectrum technic is widely used in military affairs, field such as civilian.

3 main standard WCDMA, TD-SCDMA of 3G communication, CDMA-2000 are based on spread spectrum technic.Present most widely used satellite navigation system---global positioning system (GPS) is also based on spread spectrum technic.Spread spectrum code acquisition then is a basic foundation of realizing spread spectrum communication.Only finished sign indicating number and caught and to set up synchronously at receiving terminal and transmitting terminal, could realize despreading, finished signal processing function at receiving terminal.

Short-period spreading code is widely used in modern communications, the navigation positioning system.With most widely used global position system GPS is example, and the C/A sign indicating number among the GPS is a kind of typical short-period spreading code, and bit rate is 1.023Mchip/s, and chip period length is 1023 chips, and the time cycle is 1ms.The C/A sign indicating number is the basis that GPS provides standard setting service (SPS), is widely used among the various civil navigation positioning services, for example vehicle, personal navigation service, marine vessel navigation Service, civil aviation navigation Service and mapping, entertainment service etc.

Because the relative motion between satellite and user, the carrier doppler scope is bigger, often can reach tens kHz.Rein in scope for ± 10kHz, frequency search step-length are that 1kHz calculates with many pieces of carrier waves, then have 20 frequency cells to need search.C/A code phase search scope is 1023 base sign indicating numbers, is step-length with 1/2 base sign indicating number, and it is to be searched then to have 2046 chips.So just, there are 40920 time frequency unit to need search.Be the time-frequency search two-dimensional representation of spreading code as shown in Figure 1, wherein K represents frequency cells number to be searched, and N represents code phase unit number to be searched.Existing spreading code search technique mainly contains: serial time-frequency search technique, frequency parallel search technology, code phase parallel search technology.To be introduced respectively below.

Serial search methods is with each time frequency unit lattice of correlator form serial search.As shown in Figure 1, this method is at first searched for the pairing N of first a frequency grid code phase unit in the mode of serial, and then search for pairing N code phase unit of the second lattice frequency grid, search for the 3rd the pairing N of a frequency grid code phase unit again, search is gone down so successively, till having searched for all time frequency unit lattice or having found cell greater than thresholding.Therefore the search speed of this method is very slow, now usefulness seldom, but this method but is the basis of modern various search techniques.

By the pairing a plurality of correlation integral value in each code phase unit are carried out the FFT computing, each code phase unit of parallel search pairing K frequency cells is to reach the purpose of parallel search frequency cells based on the frequency parallel search method of FFT.As shown in Figure 1, this method is parallel search code phase unit 1 a pairing K frequency cells at first, parallel search code phase unit 2 pairing K frequency cells again, search is gone down successively, till having searched for all time frequency unit lattice or having found cell greater than thresholding.Therefore compare with serial search methods, this method walks abreast on the search rate cell, and its search speed is than the fast K of serial search methods doubly in theory.

Code phase parallel search method based on FFT utilizes the FFT technology to calculate the correlation of input signal and local pseudo-intersymbol, has realized the parallel search of code phase with the FFT mode.As shown in Figure 1, this method elder generation pairing N code phase unit of first frequency cells of parallel search, and then second the pairing N of frequency cells code phase unit of parallel search, search is gone down successively, till having searched for all time frequency unit lattice or having found cell greater than thresholding.Compare with serial search methods, this method walks abreast on the searching code phase unit, therefore in theory its search speed be serial search methods N doubly.

Serial search methods all is ways of search of serial on code phase territory, frequency domain, searches for each time frequency unit lattice successively, so search speed is very slow, is difficult to satisfy the requirement of the present communication technology.

Parallel on frequency domain based on the frequency parallel search method of FFT, can parallel search frequency cells lattice.Is that 1kHz calculates with many pieces of the carrier waves scope of reining in for ± 10kHz, frequency search step-length, and the search speed of this method is 20 times of serial search methods.

Code phase parallel search method based on FFT walks abreast on time-domain, can parallel search code phase unit.With the code phase search scope is that 1023 base sign indicating numbers, step-lengths are that 1/2 basic yardage is calculated, and the search speed of this method is 2046 times of serial search methods.

By contrast, has search speed faster based on the code phase parallel search method of FFT than serial search methods with based on the frequency parallel search method of FFT.It should be noted that how many search speeds of being mentioned in above-mentioned these methods improved doubly, is to have realized and the needed time of line correlation is to obtain on the identical basis having given tacit consent to their.During actual the realization, the speed that method improves is also relevant with concrete factors such as realization hardware.

Though the code phase parallel search method based on FFT in above-mentioned three kinds of methods has the highest search speed, but this method has been subjected to very big restriction on the search rate cell.As everyone knows, in the acquiring pseudo code process, there is a pair of like this paradox: the coherent integration time T _cWith frequency cells lattice f _dBetween contradiction.The coherent integration time is long more, and the despreading gain is big more, but is subjected to the influence of frequency also big more.This influence can be used sinc ²(π f _dT _c) expression.With the coherent integration time cycle is that 1ms (a C/A sign indicating number cycle) is an example, if frequency cells lattice step-length f _Step=1000Hz is then because the caused relevant signal to noise ratio loss of carrier frequency shift can reach 3.92dB; If frequency cells lattice step-length f _Step=1600Hz is then because the caused relevant signal to noise ratio loss of carrier frequency shift can reach 12.62dB; If frequency cells lattice step-length f _Step=1800Hz is then because the caused relevant signal to noise ratio loss of carrier frequency shift can reach 19.23dB.In order to reduce loss, to have to search for less frequency step, this will make time frequency unit number to be searched sharply increase, and make search speed slower.On the other hand, if reduce the correlation intergal time, then can not realize the parallel search of code phase, search speed also can slow down, moreover can not unrestrictedly reduce the correlation intergal time.Another defective of this method is, even the signal to noise ratio of input signal is very high, its frequency search step-length can not be oversize.For example the correlation intergal time is 1ms, as frequency cells lattice step-length f _StepDuring=1800Hz, can reach 19.2320dB by the caused loss of carrier frequency shift; If frequency cells lattice step-length f _StepDuring=1900Hz, can reach 25.6dB by the caused loss of carrier frequency shift.Therefore even this method is also had to that reference carrier frequency to be searched is divided into a plurality of frequency cells lattice at signal under than very high situation and searched for, this has just limited the performance based on the code phase parallel search method of FFT.

Summary of the invention

The objective of the invention is to: realize realizing the parallel search of carrier frequency unit and code phase unit simultaneously under the condition of scale that in not obvious increase method complexity and hardware the search speed of raising method reduces the acquiring pseudo code time.

The inventor has reached the purpose of parallel search frequency cells by the received signal under the different frequency is folded; Realize the relevant of received signal and local pseudo-intersymbol by FFT, reached the purpose of parallel search code phase unit, thereby made this method possess the ability of while parallel search carrier frequency unit and code phase unit.Be " based on the short-period spreading code time frequency parallel searching method of frequency folding " schematic diagram that the inventor proposed as shown in Figure 2.

This method is determined frequency cells number to be searched according to frequency range to be searched and frequency search step-length.Then the signal on each frequency cells lattice is folded, obtain one section new signal.Utilize that the FFT mode realizes new signal and local pseudo-intersymbol and line correlation, realize the parallel search of code phase.Owing to the signal on the different frequency lattice is folded, so this method has also realized the parallel search to frequency simultaneously.Thereby this method has possessed the ability of while parallel search code phase unit and frequency cells.

This method and serial search methods, based on the frequency parallel search method of FFT, compare based on the code phase parallel search method of FFT: it has realized the parallel search of code phase unit and frequency cells simultaneously (1), rather than only realized the parallel search of code phase unit or only realized the parallel search of frequency cells, therefore having search speed faster, is a kind of complete parallel search mode truly; (2) owing to introduced the frequency uncertainty in the process of frequency folding, this method also needs " removing the frequency ambiguity degree " later on catching local pseudo-code phase---and which frequency grid is the carrier frequency of promptly determining received signal specifically drop within.Because " based on the short-period spreading code time frequency parallel searching method of frequency folding " caught pseudo-code phase earlier, therefore when search rate, only need get final product in the enterprising line frequency search of this pseudo-code phase, workload is very little.Can adopt a lot of mature methods, for example the FFT mode---the pairing a plurality of correlation integral value of this pseudo-code phase are carried out the FFT computing, to realize frequency search.Compare the whole search time of " remove the frequency ambiguity degree " needed time and method is very little, almost can ignore.

In receiver DSP signal processor, the implementation step of described " based on the short-period spreading code time frequency parallel searching method of frequency folding " is as follows:

1. at first determine frequency range f to be searched _uWith frequency search step-length f _Step, determine frequency cells number to be searched

$K=\frac{f_u}{f_{\mathrm{step}}}.$

2. receiving front-end A/D sampling KT _pThe digital signal r (n) of second (n=0,1,2 ..., KN-1).Cycle at this supposition short-period spreading code is T _p, sample frequency is f _s=N/T _pThese signals etc. are divided into K group b longways _k(n) (n=0,1,2 ..., N-1; K=0,1,2 ..., K-1), every group length is T _pSecond.Then this K group signal is mapped to respectively on the frequency cells lattice to be searched, one group of corresponding frequency grid of signal, and the signal after the mapping folded obtain new signal $b\left(n\right)=\underset{k=0}{\overset{K-1}{\mathrm{\Σ}}}{b}_k\left(n\right)$ (n＝0，1，2，...，N-1)。

The signal b (n) of doubling poststack (n=0,1,2 ..., N-1) carry out the FFT computing $B\left(m\right)=\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}b\left(n\right)e^{-j\frac{2\mathrm{\π}}{N}\mathrm{nm}},$ m＝0，1，2，...，N-1， $j=\sqrt{-1.}$

4. B (m) is got complex conjugate L as a result afterwards with local pseudo-code l (n) FFT computing ^*(m) multiply each other, and Y (m) is as a result carried out the IFFT conversion change back to time domain and obtain correlated results d (n).

$L\left(m\right)=\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}l\left(n\right)e^{-j\frac{2\mathrm{\π}}{N}\mathrm{nm}},$ m＝0，1，2，...，N-1， $j=\sqrt{-1};$

Y (m)=B (m) L ^*(m), m=0,1,2 ..., N-1, * represent to get complex conjugate

$d\left(n\right)=\frac{1}{N}\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}Y\left(m\right)e^{-j\frac{2\mathrm{\π}}{N}\mathrm{mn}},$ n＝0，1，2，...，N-1

5. store all N point ‖ d (n) ‖ ²Value (n=0,1,2 ..., N-1), " ‖ ‖ " represents delivery.The result is designated as ‖ d _i(n) ‖ ², i is a noncoherent accumulation number of times footnote.

6. repeat above-mentioned steps 2 to 5 F time altogether, carry out noncoherent accumulation then, obtain the result $c\left(n\right)=\underset{i=0}{\overset{F-1}{\mathrm{\Σ}}}{\left|\right|d_i\left(n\right)\left|\right|}^2,$ n＝0，1，2，...，N-1。

7. from c (n), select maximum $c_{\max }=\underset{n}{\max }\left\{c\left(n\right)\right\},$ And determine pseudo-code phase to be searched according to the position at maximum place.

8. carry out frequency search according to the position at pseudo-code phase place in the step 7, obtain carrier frequency.Now existing a lot of methods can realize the search to frequency, and " based on the frequency parallel search method of FFT " that for example can adopt in 3.2 to be introduced realizes the search to carrier frequency.

In the step 7 in resulting pseudo-code phase and the step 8 resulting frequency be pseudo-code phase to be searched and carrier frequency.

Have search speed faster based on the code phase parallel search method of FFT than serial search methods with based on the frequency parallel search method of FFT, therefore only just the inventor proposed compares based on the short-period spreading code time frequency parallel searching method of frequency folding and existing code phase parallel search method based on FFT at this.

As Fig. 3, Fig. 4, Fig. 5 be with GPS C/A sign indicating number be under the situation of example based on the short-period spreading code time frequency parallel searching method of frequency folding with based on the simulation performance of code phase parallel search method under varying input signal signal to noise ratio (snr) condition of FFT relatively.Simulated conditions is: through the random value between-5kHz and 5kHz of the input signal carrier frequency after the front-end processing, promptly the frequency uncertainty is ± 5kHz; Frequency search step-length f _Step=500Hz is respectively-5000Hz in the centre frequency of this each frequency search cell ,-4500Hz ... ,-500Hz, 0Hz, 500Hz ..., 5000Hz, having 21 frequency grid needs search; The code phase step-length is 1/2 base sign indicating number, and having 2046 code phases needs search; The code phase offset amount is 1000; Coherent integration length is C/A sign indicating number cycle, i.e. a 1ms; The noncoherent accumulation number of times is 1, i.e. F=1; Fig. 3 input signal SNR is-10dB that Fig. 4 input signal SNR is-5dB that Fig. 5 input signal SNR is 5dB.

F among Fig. 3 (a), Fig. 4 (a), Fig. 5 (a) _dExpression if adopt based on the code phase parallel search method of FFT then in this method all frequency grid of searching for the simulation performance of frequency input signal spacing reckling; F among Fig. 3 (b), Fig. 4 (b), Fig. 5 (b) _dExpression and corresponding Fig. 3 (a), Fig. 4 (a), the simulation performance when the pairing minimum frequency spacing of Fig. 5 (a) differs 1000Hz (i.e. 2 frequency grid).For example among Fig. 3, the frequency input signal skew is 882Hz, then the pairing minimum frequency spacing of Fig. 3 (a) is 118Hz (centre frequency be the frequency grid of 1000Hz and the gap between frequency input signal), with Fig. 3 (a) at a distance of 1000Hz be to be the frequency grid of centre frequency with 0Hz, spacing is 882Hz between this frequency grid and frequency input signal, shown in Fig. 3 (b).Fig. 3 (c), Fig. 4 (c), Fig. 5 (c) are illustrated respectively under each signal to noise ratio condition the simulation performance based on the short-period spreading code time frequency parallel searching method of frequency folding.

From Fig. 3 (c), Fig. 4 (c), Fig. 5 (c), can see, still can find correct code phase location through the short-period spreading code time frequency parallel searching method based on frequency folding behind the frequency folding.And for for the code phase parallel search method of FFT, when gap between search rate lattice and frequency input signal is big, as the 936Hz among 1155Hz, Fig. 5 (b) among 882Hz, Fig. 4 (b) among Fig. 3 (b), this method can't find correct correlation peak.Because frequency input signal is random distribution between-5kHz and 5kHz, having to that based on the code phase parallel search method of FFT frequency range to be searched is divided into a plurality of frequency grid searches for, compare with the short-period spreading code time frequency parallel searching method based on frequency folding that the inventor is proposed, its search speed will be slower.

Limitation based on the code phase parallel search method of FFT has also obtained further confirmation in Fig. 5: even promptly the input signal signal to noise ratio is higher, this method is also had at the enterprising line search of a plurality of frequency cells lattice.Shown in Fig. 5 (b), this moment the input signal signal to noise ratio up to 5dB, but when the frequency input signal skew is 936Hz, still can not find correct code phase location based on the code phase parallel search method of FFT.In fact this moment by the caused loss of the frequency shift (FS) of 936Hz up to 23.36dB, therefore having to frequency partition based on the code phase parallel search method of FFT is that a plurality of frequency cells lattice are searched for, with compare its search speed based on the short-period spreading code time frequency parallel searching method of frequency folding certainly will be slower.

Under the superincumbent simulated conditions, distribute if consider that the input signal carrier frequency shift is obeyed evenly between-5kHz and 5kHz, then the code phase parallel search method based on FFT on average needs to search for 10.5 lattice frequency cells lattice, and the short-period spreading code time frequency parallel searching method based on frequency folding that the inventor proposed then only needs disposablely can search for all frequency grid by simple frequency is folding.Promptly be based on 10.5 times of code phase parallel search method of FFT based on the search speed of the short-period spreading code time frequency parallel searching method of frequency folding.Even consider folding and solve the frequency ambiguity degree regular hour need be spent, also be based on the code phase parallel search method several times of FFT based on the search speed of the short-period spreading code time frequency parallel searching method of frequency folding.Frequency cells lattice to be searched are many more, and are all the more obvious based on the advantage of the short-period spreading code time frequency parallel searching method of frequency folding.

Description of drawings

Fig. 1. the time-frequency two-dimensional search schematic diagram of spreading code, dash area is represented time frequency unit lattice to be searched among the figure;

Fig. 2. the FB(flow block) of short-period spreading code time frequency parallel searching method based on frequency folding proposed by the invention;

Fig. 3. emulation experiment figure under the input signal SNR=-10dB: (a) f _d=118Hz, (b) f _d=882Hz is (c) behind the frequency folding;

Fig. 4. emulation experiment figure under the input signal SNR=-5dB: (a) f _d=155Hz, (b) f _d=1155Hz is (c) behind the frequency folding;

Fig. 5. emulation experiment figure under the input signal SNR=5dB: (a) f _d=64Hz, (b) f _d=936Hz is (c) behind the frequency folding.

Embodiment

Is that example is told about the implementation step based on the short-period spreading code time frequency parallel searching method of frequency folding at this with GPS C/A sign indicating number, and the enforcement of method is not limited to GPS C/A sign indicating number and following these concrete parameters, can select flexibly.GPS C/A sign indicating number is a kind of typical short-period spreading code, and bit rate is 1.023Mchip/s, and chip period length is 1023 chips, and the time cycle is 1ms.Suppose that the code phase step-length is 1/2 base sign indicating number, frequency range to be searched be ± 5kHz, and the frequency search step-length is 500Hz, 2046 code phases, 20 frequency cells lattice then arranged, amount to 40920 time frequency unit and need to search for.The noncoherent accumulation number of times is assumed to 10 times.In the above conditions, the implementation step based on the short-period spreading code time frequency parallel searching method of frequency folding is as follows:

1. select 40920 signals after front-end A/D sampling, be divided into 20 groups, every group length is 2046 points.Every group of signal is mapped to different frequency cells lattice respectively, and the corresponding addition of the signal after will shining upon then obtains 2046 new signals, and these 2046 signals is carried out the FFT computing;

2. produce one-period and be the local C/A sign indicating number of 1ms and it is sampled, obtain 2046 local pseudo-codes and these 2046 signals are carried out the FFT computing and get complex conjugate.This step can be finished in advance, all repeats during lock-on signal at every turn;

3. 1 and 2 result is multiplied each other and carry out IFFT and handle;

4. store 2046 resulting results of step 3;

5. repeat above-mentioned steps 1 to 4 (wherein step 2 need are carried out and once the result stored, and needn't repeat) 10 times;

6. 10 results with step 5 gained carry out noncoherent accumulation, and carry out power detection, therefrom find out maximum (or finding out than the big person of thresholding);

7. determine pseudo-code phase according to 6 gained maximums (or greater than thresholding person);

8. according to 7 gained pseudo-code phase,, solve the frequency ambiguity degree in the enterprising line frequency search of this pseudo-code phase.A kind of method commonly used is, obtains 10 groups of coherent integration values on this pseudo-code phase, carries out the FFT conversion, determines the carrier frequency shift of received signal according to the position at gained frequency spectrum maximum place.

Illustrate: (1) in order to improve FFT efficient, and the mode that can adopt zero padding of counting of the FFT in the step 1,2 is supplemented to 2048 points; (2) a kind of way of definite thresholding in the step 6: establish P _FaThe false alarm probability of the single unit to be searched that expression is set, δ ²The expression noise variance, A is the received signal amplitude, and K represents frequency folding number of times (being 20) in this example, and N represents sampling number (being 2046) in the one-period in this example, and then thresholding can be set at $P_o=\sqrt{-2\mathrm{KN}({\mathrm{\δ}}^2+A^2)\ln P_{\mathrm{fa}}}.$

Claims (2)

1. based on the short-period spreading code time frequency parallel searching method of frequency folding, it is characterized in that this method is to realize as follows in the DSP of receiving terminal signal processor:

Step (1) initialization

To DSP signal processor input frequency range f to be searched _uWith frequency search step-length f _Step, and frequency cells to be searched is counted K=f _u/ f _Step

This DSP of step (2) receives the KT that samples through front-end A/D _pThe digital signal r (n) of second, n=0,1,2 ..., KN-1.T _pBe the cycle of short-period spreading code, set point; R (n) etc. is divided into K group signal b longways _k(n), n=0,1,2 ..., N-1, k=0,1,2 ..., K-1, every group length is T _pSecond.And then K group signal b _k(n) be mapped to respectively on the frequency cells lattice to be searched, make the corresponding frequency grid of one group of signal, and the signal after the mapping is folded, obtain signal b (n):

$b\left(n\right)=\underset{k=0}{\overset{K-1}{\mathrm{\Σ}}}{b}_k\left(n\right),$ n＝0，1，2，...，N-1；

FFT unit among this DSP of step (3) to step (2) obtain folding after signal b (n) carry out the FFT computing, obtain signal B (m):

$B\left(m\right)=\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}b\left(n\right)e^{-j\frac{2\mathrm{\π}}{N}\mathrm{nm}},$ m＝0，1，2，...，N-1， $j=\sqrt{-1};$

Signal B (m) and signal L that step (4) obtains step (3) ^*(m) multiply each other, obtain signal Y (m), again this signal Y (m) is carried out the IFFT conversion, obtain time-domain signal d (n):

$L\left(m\right)=\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}l\left(n\right)e^{-j\frac{2\mathrm{\π}}{N}\mathrm{nm}},$ m＝0，1，2，...，N-1， $j=\sqrt{-1},$ L (m) is local pseudo-code l (n) FFT calculated result,

Y (m)=B (m) L ^*(m), m=0,1,2 ..., N-1, L ^*(m) be the complex conjugate of signal L (m),

$d\left(n\right)=\frac{1}{N}\underset{m=0}{\overset{N-1}{\mathrm{\Σ}}}Y\left(m\right)e^{j\frac{2\mathrm{\π}}{N}\mathrm{mn}},$ n＝0，1，2，...，N-1；

Step (5) is stored all N point ‖ d (n) ‖ ²Value, n=0,1,2 ..., N-1, " ‖ ‖ " expression delivery symbol;

Step (6) repeating step (2)～step (5), F time altogether, the result that each step (5) is obtained is designated as ‖ d ₁(n) ‖ ², i is a noncoherent accumulation number of times footnote, carries out noncoherent accumulation then, obtains signal c (n):

$c\left(n\right)=\underset{i=0}{\overset{F-1}{\mathrm{\Σ}}}{\left|\right|d_i\left(n\right)\left|\right|}^2,$ n＝0，1，2，...，N-1，i＝1，2，...，F，

Wherein F represents the number of times according to noncoherent accumulation, is recently determined according to the noise of received signal by the user;

Step (7) is selected maximum c from c (n) _Max:

$c_{\max }=\underset{n}{\max }\left\{c\left(n\right)\right\},$

According to maximum c _MaxPseudo-code phase to be searched is determined in the position at place;

Step (8) is used based on the frequency parallel search method of FFT and is searched in the enterprising line frequency of the pseudo-code phase of step (7) gained, obtains carrier frequency;

In step (7), can be to the result behind the noncoherent accumulation with setting thresholding P _oMethod replace looking for maximum c _Max:

$P_o=\sqrt{-2\mathrm{KN}({\mathrm{\δ}}^2+A^2)\ln P_{\mathrm{fa}}},$

P _FaBe the false alarm probability of the single unit to be searched set of user, δ ²The expression noise variance, A is the received signal amplitude, and K represents the frequency folding number of times, and N is by the sampling number decision of short-period spreading code in one-period.

2. the short-period spreading code time frequency parallel searching method based on frequency folding according to claim 1 is characterized in that, in step (2) signal on the different frequency cell has been carried out K time and has folded.

Images