CN102480452A - Carrier frequency synchronization circuit and method of OFDM system - Google Patents

Abstract

The invention discloses a carrier frequency synchronization circuit and a method of an OFDM system, which comprises circuits such as decimal frequency offset estimation, decimal frequency offset compensation, integer frequency offset estimation, integer frequency offset compensation and the like; correspondingly, the method comprises the steps of fractional frequency offset estimation, fractional frequency offset compensation, integer frequency offset estimation and integer frequency offset compensation; performing normalization processing on cross-correlation values of two under-sampled synchronous sequences adjacent to a frequency domain to obtain an estimated value of fractional frequency offset, and performing fractional frequency offset compensation correction on the estimated value; then, carrying out autocorrelation operation on the high-energy carrier sign bit of each OFDM symbol after Fast Fourier Transform (FFT) processing, carrying out summation operation on the autocorrelation of a plurality of OFDM symbols to obtain an integer frequency offset estimation value, and then carrying out integer frequency offset compensation according to the integer frequency offset estimation value. By adopting the synchronous circuit and the method, the operation amount can be reduced, the influence of noise on frequency offset estimation can be reduced, and the design and the realization of low-power consumption equipment are facilitated.

Description

A kind of carrier frequency synchronization circuit and method of ofdm system

Technical field

The present invention relates to the signal of communication modulation technique, relate in particular to the carrier frequency synchronization circuit and the method for a kind of OFDM (OFDM) system.

Background technology

The OFDM technology disturbs because of its anti-multipath and the ability of anti-frequency selective fading is widely used in the multiple high-speed data access system, like occasions such as WLAN, high-speed digital subscriber line, ADSL, digital audio broadcasting, DVB and HD digital TVs.Present many worlds, domestic standard are with the transmission means of OFDM technology as physical layer.

Because the OFDM technology can solve multipath in the high-speed communication system preferably and disturb and problem such as wideband transmit, therefore should technology have become the candidate scheme of the 4th third-generation mobile communication technological (B3G/4G).Standards such as satellite broadcasting at present, digital moving-tv broadcasting, DTB Digital Terrestrial Broadcasting all are to adopt the OFDM technology as core technology.But but have a problem: because the OFDM technology relies on the orthogonal property of frequency, in case its orthogonality is destroyed, the error rate of its system will sharply increase, and this also is the major defect of OFDM technology in application.Its main cause that produces frequency deviation is that radio circuit receives and delivery section centre frequency mismatch; Another reason is that receiving-transmitting sides has the Doppler frequency shift of high speed relative velocity.

In ofdm system, when producing carrier frequency offset, partly be called integer frequency bias Δ f greater than the frequency deviation of subcarrier spacing _I, integer frequency bias just is shifted the output of fast Fourier transform (FFT) in the receiver, does not destroy the orthogonality between each number of sub-carrier, but will cause the complete mistake of demodulation result; Frequency deviation less than subcarrier spacing partly is called decimal frequency bias Δ f _f, the orthogonality that it can destroy subcarrier causes subcarrier interference, causes system's error code to rise.Generally, need at first to carry out that decimal frequency bias is estimated and compensation in time domain, with eliminate disturb between the subcarrier that decimal frequency bias causes after, carry out at frequency domain again that integer frequency bias is estimated and compensation.

In radio data system, because frequency spectrum resource is of great rarity, system bandwidth is generally all narrow.In addition, because the channel circumstance very severe of long Distance Transmission, system must resist bigger time delay expansion, so the system subcarrier number is generally many, thereby each subcarrier spacing is all smaller.Like this, system recipient's carrier frequency offset is generally all greater than subcarrier spacing, need be divided into the integer carrier wave frequency deviation and the decimal carrier wave frequency deviation is estimated respectively to system.Wherein:

For decimal frequency bias; Mainly be to utilize correlation properties to estimate, as based on time-delay correlation technique of targeting sequencing etc., the subject matter of this method is that operand is big, power consumption is high; And portable terminal is very responsive to power consumption, and therefore designing requirement height, the difficulty to low power consuming devices is big.

For integer frequency bias, can be utilized in the scattered pilot or the CP continuous pilot that insert in the frequency domain and carry out integer frequency bias Δ f _IEstimate that it is theoretical based on maximum likelihood that integer frequency bias is estimated to be actually.Because decimal frequency bias is proofreaied and correct, the interference between the subcarrier is eliminated basically.Originally the k ' th number of sub-carrier is because integer frequency bias has been displaced to k ' th+ Δ f after the FFT demodulation process _ISet up departments the system in CP continuous pilot c _{K '}Its sub-carrier positions k ' ∈ C, the CP continuous pilot after demodulation that then receives moves to k ∈ C+ Δ f _ITo two continuous OFDM symbol l-1, l carries out conjugation and is correlated with, and asking the position of its maximal peak point skew ideal point is integer frequency bias.But the shortcoming of said method is that the influence that this estimation procedure is disturbed by noise and multipath fading phenomenon etc. is bigger.Especially CP continuous pilot generally sends with burst of energy, and correlation peak point is more obvious like this, like document Speth, M.; Fechtel, S.; Fock, G.; Meyr, H., Optimum receiver design for OFDM-based broadband transmission.II.A case study, IEEE Transactions on Communications, Vol.49, Apr.2001, formula 12 is (promptly in PP571～578 X wherein _kBe the conjugation correlation of two adjacent OFDM symbol CP continuous pilot, C is the sub-carrier positions of CP continuous pilot, and m is the scope that CP continuous pilot horizontally slips) described content.

In addition, can also adopt based on the protection section of inserting and carry out the integer Nonlinear Transformation in Frequency Offset Estimation.In ofdm system, data subcarrier and pilot frequency carrier wave sum are generally all less than the FFT length N, and remaining part realizes through filling the virtual subnet carrier wave.Just be equivalent to have a protection section like this at the frequency spectrum two ends, in order to reduce the interference with successive bands.Here, in each OFDM symbol, have the stator carrier wave not load any information, its transmitted power is 0, and this section virtual subnet carrier wave is called the protection section.Theoretical based on maximum likelihood, when data and pilot sub-carrier not in the sliding window of design in advance, the energy in the sliding window reaches minimum value, the distance between sliding window and the ideal position is exactly the size of integer frequency bias at this moment.It is bigger that this algorithm is influenced by sliding window width and deep fade, need between complexity that the sliding window width is caused and performance, do compromise.

Except that said method; Can also adopt that the patent No. is 03100300.1, name is called in the Chinese patent of " in DVB, utilizing time-domain spread spectrum beacon to carry out synchronous method " the integer Nonlinear Transformation in Frequency Offset Estimation method based on targeting sequencing; Promptly utilize the targeting sequencing that receives after the time synchronized; After decimal frequency bias is proofreaied and correct, multiply by [k one by one _Max, k _Max-1] the integer carrier wave frequency deviation in the scope carries out computing cross-correlation with desirable targeting sequencing respectively then, and making this correlation obtain peaked i is exactly integer frequency bias.

In sum; Decimal, integer frequency bias method of estimation in the existing ofdm system carrier frequency synchronization method; The influence that exists estimation procedure disturbed by noise and multipath fading phenomenon etc. is bigger, or amount of calculation is big, is not suitable for being applied in the defective in the equipment that requires low complex degree calculating.

Summary of the invention

Main purpose of the present invention is to provide a kind of carrier frequency synchronization circuit and method of ofdm system; Utilize this synchronous circuit; Can under the prerequisite that guarantees systematic function, reduce algorithm complex and hardware spending, be beneficial to the design of low-power consumption portable terminal.

For achieving the above object, technical scheme of the present invention is achieved in that

A kind of carrier frequency synchronization circuit of orthogonal frequency division multiplex OFDM system comprises decimal frequency bias estimating circuit, decimal frequency bias compensating circuit, integer frequency bias estimating circuit and integer frequency bias compensating circuit; Wherein,

Said decimal frequency bias estimating circuit is used for two adjacent cross correlation values of owing the sample-synchronous sequence of frequency domain are carried out the normalization processing, and obtains the decimal frequency bias estimated value;

Said decimal frequency bias compensating circuit is used for carrying out the decimal frequency bias compensation according to said decimal frequency bias estimated value input digital controlled oscillator, to proofread and correct the carrier frequency of said ofdm system;

Said integer frequency bias estimating circuit is used for each the OFDM symbol high-energy carrier wave sign bit after handling through FFT FFT is carried out auto-correlation computation, and the autocorrelation value of a plurality of OFDM symbols is carried out summation operation, obtains the integer frequency bias estimated value;

Said integer frequency bias compensating circuit is used for carrying out the integer frequency bias compensation according to said integer frequency bias estimated value digital controlled oscillator, to proofread and correct the carrier frequency of said ofdm system.

A kind of carrier frequency synchronization method of ofdm system; After said ofdm system time synchronized, handle the phase angle of peaking point through the autocorrelation value of adjacent synchronizing sequence being carried out normalization; Obtain the decimal frequency bias estimated value, and the carrier frequency of ofdm system is carried out the decimal frequency bias compensation; Again will be further after FFT FFT handles through the data after the decimal frequency bias compensation; Each OFDM symbol high-energy carrier wave sign bit is carried out auto-correlation computation; And the auto-correlation of a plurality of OFDM symbols is carried out summation operation carry out the integer frequency bias estimation; Obtain the integer frequency bias estimated value, compensate according to this integer frequency bias estimated value at last, to proofread and correct the carrier frequency of ofdm system.

Wherein, described decimal frequency bias is estimated to comprise:

The cross correlation value that sample rate is adjacent two synchronizing sequences of α is owed in A, calculating:

$P_{\mathrm{xcorr}}\left(n\right)=\underset{k=0}{\overset{N-1}{\mathrm{\Σ}}}{r}_{n+\mathrm{\αk}}\mathrm{conj}\left(r_{n+\mathrm{\αk}+N}\right)$

Wherein, α is for owing sampled value, r _nBe n the sampled value that receives, N is a synchronizing sequence length, and conj (.) is a conjugate operation.

The autocorrelation value that sample rate is adjacent two synchronizing sequences of α is owed in B, calculating:

$P_{\mathrm{auto}}\left(n\right)=\underset{k=0}{\overset{N-1}{\mathrm{\Σ}}}{r}_{n+\mathrm{\αk}+N}\mathrm{conj}\left(r_{n+\mathrm{\αk}+N}\right)$

Wherein, α is for owing sampled value, r _nBe n the sampled value that receives, N is a synchronizing sequence length, and conj (.) is a conjugate operation.

The peak point phase angle of C, calculating normalized crosscorrelation value, i.e. decimal frequency bias estimated value:

$\mathrm{\Δ}{f}_f=\frac{f_s}{2\mathrm{\πN}}\arg \left(\max \left(\frac{P_{\mathrm{xcorr}}\left(n\right)}{P_{\mathrm{auto}}\left(n\right)}\right)\right)$

Wherein, fs is a sample frequency, and N is a synchronizing sequence length, and max (.) is the maximizing computing, and arg (.) is for asking plural phase angle computing.

Described integer frequency bias estimation procedure comprises:

D, each OFDM symbol high-energy carrier wave sign bit of frequency domain is carried out auto-correlation computation:

z _l，k+i＝sign(real(Z _l，k+i))+j*sign(imag(Z _l，k+i))

Wherein,

Be the computing of high-energy carrier wave sign bit, shown in a represent high-energy carrier wave sign bit, shown in Th be the threshold value of dynamic adjustable; Z _{L, k+i}The frequency domain data subcarrier;

E, the autocorrelation value of a plurality of OFDM symbols is carried out summation operation, asking its minimum value is the integer frequency bias value:

Wherein, k _MinAnd k _MaxBe respectively in each OFDM symbol the sequence number of first and last virtual subnet carrier wave.

Described compensate of frequency deviation process is sent into the compensate of frequency deviation module for the decimal frequency bias that calculates step C or step e gained and integer frequency bias and is compensated:

$r=r_n{e}^{-j2\mathrm{\πn}\frac{\mathrm{\Δ}{f}_f+\mathrm{\Δ}{f}_I}{f_s}}$

Wherein, r be receive data through the result of compensate of frequency deviation, r _nBe n the sampled value that receives, fs is a sample frequency.

The carrier frequency synchronization circuit and the method for ofdm system provided by the present invention have the following advantages:

Through two adjacent synchronizing sequences of frequency domain being owed sampling, then its cross correlation value is carried out normalization and handle, obtain the estimated value of decimal frequency bias, reduced the amount of calculation that decimal frequency bias is estimated to a great extent.Through each the OFDM symbol high-energy carrier wave sign bit after handling through fast Fourier transform (FFT) is carried out auto-correlation computation, obtain the side-play amount of its minimum value and ideal value, and then obtain the integer frequency bias estimated value, to reduce hardware spending.Auto-correlation through to a plurality of OFDM symbols is carried out summation operation, reduces channel estimation errors to the integer frequency bias estimation effect.In addition, the threshold value of high-energy carrier wave is chosen, can self adaptation be adjusted to reduce the influence of system noise to systematic function according to the difference of system works environment.

Description of drawings

Fig. 1 is a carrier frequency synchronization electrical block diagram of the present invention;

Fig. 2 is the carrier frequency synchronization method flow diagram of ofdm system of the present invention.

Embodiment

Below in conjunction with accompanying drawing and embodiments of the invention method of the present invention is done further detailed explanation.

Fig. 1 is a carrier frequency synchronization electrical block diagram of the present invention; As shown in Figure 1; Comprise decimal frequency bias estimating circuit, decimal frequency bias compensating circuit, integer frequency bias estimating circuit and integer frequency bias compensating circuit; Wherein, said decimal frequency bias compensating circuit, integer frequency bias compensating circuit are claimed the frequency compensation module again.

The present invention utilizes this carrier frequency synchronization circuit, at first must carry out the normalization processing to two adjacent cross correlation values of owing the sample-synchronous sequence of frequency domain, and the phase angle of peaking point promptly carries out decimal frequency bias and estimates, obtains the estimated value of decimal frequency bias; Again said decimal frequency bias estimated value input digital controlled oscillator (NCO) is carried out the decimal frequency bias compensation,, it is kept synchronously to proofread and correct the carrier frequency of ofdm system.Then each the OFDM symbol high-energy carrier wave sign bit after handling through fast Fourier transform (FFT) is carried out auto-correlation computation, promptly carry out integer frequency bias and estimate, again the autocorrelation value of a plurality of OFDM symbols is carried out summation operation, obtain the integer frequency bias estimated value; At last, more said integer frequency bias estimated value being sent into the compensate of frequency deviation module proofreaies and correct the carrier frequency of ofdm system.

As a preferred embodiment, suppose that in signal bandwidth be in the wireless broadcast system of 10M, subcarrier numerical digit 4096, highest frequency is 2.6GHz in the wavelength coverage of ofdm system place, subcarrier spacing is 10MHz/4096=2.44KHz.Requiring the frequency deviation of maximum tolerance for transmit leg and recipient also is 20ppm, so the frequency offset estimation range of frequency synchronisation should be 40ppm.So its frequency deviation is 2.6GHz * 40ppm=104KHz, this frequency deviation is equivalent to the 104KHz/2.44KHz=42.6 number of sub-carrier at interval.

From finding out here, because system's sub-carriers is very little at interval, the skew that is caused by frequency deviation is very big, and maximum can squint near 43 number of sub-carrier.So need frequency deviation be divided into decimal and the integral multiple subcarrier spacing estimates respectively.

Fig. 2 is the carrier frequency synchronization method flow diagram of ofdm system of the present invention, and as shown in Figure 2, this carrier frequency synchronization method comprises the steps:

Step 201, calculating decimal frequency bias estimated value.Specific as follows:

At first, utilize formula (1) to calculate and owe the cross correlation value P that sample rate is adjacent two synchronizing sequences of α _Xcorr(n):

$P_{\mathrm{xcorr}}\left(n\right)=\underset{k=0}{\overset{N-1}{\mathrm{\Σ}}}{r}_{n+\mathrm{\αk}}\mathrm{conj}\left(r_{n+\mathrm{\αk}+N}\right)---\left(1\right)$

Wherein, α is for owing sampled value; r _nBe n the sampled value that receives; K is for receiving the index after signal is owed to sample, and its span is 0,1 ..., N-1, N are the length of synchronizing sequence, conj (.) is a conjugate operation.

Then, utilize formula (2) to calculate and owe the autocorrelation value P that sample rate is adjacent two synchronizing sequences of α _Auto(n):

$P_{\mathrm{auto}}\left(n\right)=\underset{k=0}{\overset{N-1}{\mathrm{\Σ}}}{r}_{n+\mathrm{\αk}+N}\mathrm{conj}\left(r_{n+\mathrm{\αk}+N}\right)---\left(2\right)$

At last, the peak point phase angle that utilizes formula (3) to calculate the normalized crosscorrelation value is decimal frequency bias estimated value Δ f _f:

$\mathrm{\Δ}{f}_f=\frac{f_s}{2\mathrm{\πN}}\arg \left(\max \left(\frac{P_{\mathrm{xcorr}}\left(n\right)}{P_{\mathrm{auto}}\left(n\right)}\right)\right)---\left(3\right)$

F wherein _sBe sample frequency.

Step 202, carry out decimal frequency bias compensation.Utilize formula (4) to calculate:

$r=r_n{e}^{-j2\mathrm{\πn}\frac{\mathrm{\Δ}{f}_f}{f_s}}---\left(4\right)$

Wherein, r represents the result after decimal frequency bias compensates; r _nN the sampled value that representative receives;

Represent the ratio of decimal frequency bias estimated value and sample frequency.

In ofdm system, when the time-delay related amplitude value reached maximum, its peak value angular frequency was exactly a decimal frequency bias.Therefore input sample point is through owing the sampling of owing that the rate of adopting is α, carry out and memory (RAM) capacity that adds up can be reduced to 1/ original α, and operating frequency also can be reduced to 1/ α.

In the decimal frequency bias method for synchronous in this example, owing to adopt rate α is 4, and also promptly extracting counts, and to account for all ratios of counting be 1/4, and this moment, the standard variance of decimal frequency bias evaluated error was 0.022.In the method, input sample point is every at a distance from 4 somes extractions once, is used for and the RAM that adds up can be reduced to original 1/4; Operating frequency also is reduced to 1/4; Therefore, reduced the requirement to hardware, promptly the hardware implementation complexity can both obviously be reduced with the realization power consumption.

Step 203, carry out integer frequency bias and estimate, each OFDM symbol high-energy carrier wave sign bit of frequency domain is carried out auto-correlation computation.

In ofdm system, because subcarrier spacing is smaller, so integer frequency bias is bigger, its frequency deviation can reach 42 number of sub-carrier at interval.In the present embodiment, the data after the mapping are moved on the input of invert fast fourier transformation (IFFT) by formula (5).

$Z\left(i\right)=\left\{\begin{array}{cc}Y(i-1)& i\∈\left[\mathrm{1,1538}\right]\\ Y(i-1020)& i\∈\left[\mathrm{2558,4095}\right]\\ 0& i\∈\left[\mathrm{1539,2557}\right],0\end{array}\right.---\left(5\right)$

Wherein, Z (i) represents the input of IFFT and the output of mapping respectively with Y (i).

Can find out that subcarrier 1539-2557 totally 1019 number of sub-carrier realizes through filling 0, calls the virtual subnet carrier wave to these subcarriers.

Define symbol function sign () is:

$\mathrm{sign}\left(a\right)=\left\{\begin{array}{cc}1& a\&GreaterEqual;\mathrm{Th},\\ -1& a\&le;-\mathrm{Th},\\ 0& \left|a\right|<\mathrm{Th},\end{array}\right.$

Wherein, a is a representative high-energy carrier wave sign bit, and Th is the dynamic adjustable threshold value, adjusts according to the needed channel condition information of system's soft-decision;

Definition z _{L, k+i}For:

z _l，k+i＝sign(real(Z _l，k+i))+j*sign(imag(Z _l，k+i))

Wherein, Z _{L, k+i}It is the k+i number of sub-carrier of l OFDM symbol.

Then the autocorrelation value of a plurality of OFDM symbols is carried out summation operation, ask its minimum value to be the integer frequency bias value, formula capable of using (6) is tried to achieve the integer frequency bias value among the present invention:

In the ofdm system of present embodiment, each time slot is made up of 53 OFDM symbols, i.e. N _s=53.Wherein, k _MinAnd k _MaxIt is the sequence number of first and last virtual subnet carrier wave in each OFDM symbol.

Step 204, according to the integer frequency bias estimated value of step 203 gained, the sync carrier frequency of ofdm system is carried out the integer frequency bias compensation.Formula capable of using (7) calculates among the present invention, obtains the integer frequency bias compensation result:

$r=r_n{e}^{-j2\mathrm{\&pi;n}\frac{\mathrm{\&Delta;}{f}_I}{f_s}}---\left(7\right)$

Wherein, wherein, r represents the result after the integer frequency bias compensation; r _nN the sampled value that representative receives;

The ratio of integer frequency bias estimated value and sample frequency.

In the integer frequency bias method for synchronous in this example; With Chinese patent number is that the integer Algorithm of Carrier Frequency Offset Estimation that relates in 03100300.1 1 literary compositions is compared; Need plural related operation in the former method for synchronous; Only need the related operation of 1 bit (bit) to get final product and use the inventive method, computational complexity reduces greatly, and the carrier frequency bias estimation based on ofdm system promptly of the present invention has significantly reduced operand and hardware implementation complexity.

In addition, use method provided by the invention, also can reduce the influence of channel estimation errors whole frequency offset estimating.Also have, the threshold value of choosing for the high-energy carrier wave can self adaptation adjustment reduce system noise to Effect on Performance according to the difference of system works environment.This method can be widely used in various wireless broadcast systems, the communications field.

The above is merely preferred embodiment of the present invention, is not to be used to limit protection scope of the present invention.

Claims (5)

1. the carrier frequency synchronization circuit of an orthogonal frequency division multiplex OFDM system is characterized in that, comprises decimal frequency bias estimating circuit, decimal frequency bias compensating circuit, integer frequency bias estimating circuit and integer frequency bias compensating circuit; Wherein,

Said decimal frequency bias estimating circuit is used for two adjacent cross correlation values of owing the sample-synchronous sequence of frequency domain are carried out the normalization processing, and obtains the decimal frequency bias estimated value;

Said decimal frequency bias compensating circuit is used for carrying out the decimal frequency bias compensation according to said decimal frequency bias estimated value input digital controlled oscillator, to proofread and correct the carrier frequency of said ofdm system;

Said integer frequency bias estimating circuit is used for each the OFDM symbol high-energy carrier wave sign bit after handling through FFT FFT is carried out auto-correlation computation, and the autocorrelation value of a plurality of OFDM symbols is carried out summation operation, obtains the integer frequency bias estimated value;

Said integer frequency bias compensating circuit is used for carrying out the integer frequency bias compensation according to said integer frequency bias estimated value digital controlled oscillator, to proofread and correct the carrier frequency of said ofdm system.

2. the carrier frequency synchronization method of an ofdm system; It is characterized in that; After said ofdm system time synchronized, handle the phase angle of peaking point through the autocorrelation value of adjacent synchronizing sequence being carried out normalization; Obtain the decimal frequency bias estimated value, and the carrier frequency of ofdm system is carried out the decimal frequency bias compensation; Again will be further after FFT FFT handles through the data after the decimal frequency bias compensation; Each OFDM symbol high-energy carrier wave sign bit is carried out auto-correlation computation; And the auto-correlation of a plurality of OFDM symbols is carried out summation operation carry out the integer frequency bias estimation; Obtain the integer frequency bias estimated value, compensate according to this integer frequency bias estimated value at last, to proofread and correct the carrier frequency of ofdm system.

3. the carrier frequency synchronization method of ofdm system according to claim 1 is characterized in that, the computational process of described decimal frequency bias estimated value is:

The cross correlation value that sample rate is adjacent two synchronizing sequences of α is owed in A, calculating:

$P_{\mathrm{xcorr}}\left(n\right)=\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{r}_{n+\mathrm{\&alpha;k}}\mathrm{conj}\left(r_{n+\mathrm{\&alpha;k}+N}\right)$

Wherein, α is for owing sampled value, r _nBe n the sampled value that receives, N is a synchronizing sequence length, and conj (.) is a conjugate operation.

The autocorrelation value that sample rate is adjacent two synchronizing sequences of α is owed in B, calculating:

$P_{\mathrm{auto}}\left(n\right)=\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{r}_{n+\mathrm{\&alpha;k}+N}\mathrm{conj}\left(r_{n+\mathrm{\&alpha;k}+N}\right)$

Wherein, α is for owing sampled value, r _nBe n the sampled value that receives, N is a synchronizing sequence length, and conj (.) is a conjugate operation.

The peak point phase angle of C, calculating normalized crosscorrelation value, i.e. decimal frequency bias estimated value:

$\mathrm{\&Delta;}{f}_f=\frac{f_s}{2\mathrm{\&pi;N}}\arg \left(\max \left(\frac{P_{\mathrm{xcorr}}\left(n\right)}{P_{\mathrm{auto}}\left(n\right)}\right)\right)$

Wherein, fs is a sample frequency, and N is a synchronizing sequence length, and max (.) is the maximizing computing, and arg (.) is for asking plural phase angle computing.

4. the carrier frequency synchronization method of ofdm system according to claim 1 is characterized in that, the computational process of described integer frequency bias estimated value comprises:

D, each OFDM symbol high-energy carrier wave sign bit of frequency domain is carried out auto-correlation computation:

z _l，k+i＝sign(real(Z _l，k+i))+j*sign(imag(Z _l，k+i))

Wherein,

Be the computing of high-energy carrier wave sign bit, shown in a represent high-energy carrier wave sign bit, shown in Th be the threshold value of dynamic adjustable; Z _{L, k+i}The frequency domain data subcarrier;

E, the autocorrelation value of a plurality of OFDM symbols is carried out summation operation, asking its minimum value is the integer frequency bias value:

Wherein, k _MinAnd k _MaxBe respectively in each OFDM symbol the sequence number of first and last virtual subnet carrier wave.

5. according to the carrier frequency synchronization method of claim 3 or 4 described ofdm systems, it is characterized in that described compensate of frequency deviation process is sent into the compensate of frequency deviation module for the decimal frequency bias that calculates step C or step e gained and integer frequency bias and compensated:

$r=r_n{e}^{-j2\mathrm{\&pi;n}\frac{\mathrm{\&Delta;}{f}_f+\mathrm{\&Delta;}{f}_I}{f_s}}$

Wherein, r be receive data through the result of compensate of frequency deviation, r _nBe n the sampled value that receives, fs is a sample frequency.

Images