CN103001914B - Decimal based on Cyclic Prefix times Carrier Synchronization Algorithm - Google Patents

Abstract

The invention discloses a kind of times Carrier Synchronization Algorithm of the decimal based on Cyclic Prefix, solved the Carrier Synchronization accuracy existing in prior art not high enough, can not meet the problem of order of modulation growth requirement. Be somebody's turn to do the decimal times Carrier Synchronization Algorithm based on Cyclic Prefix, comprise the following steps: the signal receiving is divided into carrier wave frequency deviation part and remaining part, and carrier wave frequency deviation branch is divided into fractional part of frequency offset and integer frequency offset, and calculate the single estimated value of fractional part of frequency offsetAccording to the single estimated value of fractional part of frequency offsetObtain its estimated value of the i time; Repeated execution of steps (b), calculates the variance of gained fractional part of frequency offset, and calculates mean of variance; Thresholding V is set, if V_f>V continues execution step (a), if V_f≤ V locks current fractional part of frequency offset estimated value, and carries out compensate of frequency deviation with this estimated value all the time. By such scheme, the present invention has reached the residual error that reduces decimal overtones band and estimate, has improved the stability of a system, allows the object of system works in less residual error environment, has very high practical value and promotional value.

Description

Decimal based on Cyclic Prefix times Carrier Synchronization Algorithm

Technical field

The present invention relates to a kind of Carrier Synchronization Algorithm, specifically, relate in a kind of high-order wire communication based on followingThe decimal times Carrier Synchronization Algorithm of ring prefix.

Background technology

In communication system, synchronized algorithm is the basis that communication control processor is realized, and is the research of any modulation-demodulation techniqueEmphasis and difficult point, its performance is directly connected to the performance of whole communication system. Can say that synchronized algorithm is to realize information reliably to passDefeated prerequisite, only has synchronized algorithm accurately just can carry out reliable transfer of data.

At present, ofdm system synchronously can be divided into frame synchronization, sign synchronization, carrier synchronization and sampling clock synchronous. ItsIn, carrier synchronization is used for estimating the frequency shift (FS) of receiving end signal, and it is corrected to existing OFDM Carrier SynchronizationResidual error can only be controlled to below 1% subcarrier spacing, thereby can only support 16～64AM order of modulation, but, HINOC forImprove data transfer rate and order of modulation has been brought up to 16～1024QAM, existing Carrier Synchronization obviously can not meet itDemand, we need one frequency deviation residual error can be reduced to 1e^-3Even 1e^-4The synchronized algorithm of subcarrier spacing rank.

Summary of the invention

The object of the present invention is to provide a kind of times Carrier Synchronization Algorithm of the decimal based on Cyclic Prefix, mainly solve existingThe Carrier Synchronization accuracy existing in technology is not high enough, can not meet the problem of order of modulation growth requirement.

To achieve these goals, the technical solution used in the present invention is as follows:

Decimal based on Cyclic Prefix times Carrier Synchronization Algorithm, comprises the following steps:

(a) signal receiving is divided into carrier wave frequency deviation part and remaining part, and carrier wave frequency deviation branch is divided into littleSeveral times frequency deviation and integer frequency offset, and calculate the single estimated value of fractional part of frequency offset

(b) according to the single estimated value of fractional part of frequency offsetObtain its estimated value of the i time, wherein, i is positive integer;

(c) repeated execution of steps (b), calculates the variance of gained fractional part of frequency offset, and calculates mean of variance V_f；

(d) thresholding V is set, if V_f> V continues execution step (a), if V_f≤ V locks current fractional part of frequency offset and estimatesEvaluation, and carry out compensate of frequency deviation with this estimated value all the time.

Specifically, in described step (a), the single estimated value of decimal times carrier wave frequency deviationDrawn by following steps:

(a1) k the time domain sampling point r (k) that establishes the OFDM symbol receiving after Complete Synchronization is: r (k)=s (k) exp(j2πkf_F/N)+n(k)；

(a2) summation of the related operation based on cyclic prefix CP drawsWherein N isData segment, length 256, N_CPFor circulating prefix-length 16,The OFDM symbol original position obtaining for sign synchronization; And then drawThe single estimated value of fractional part of frequency offset

Further, in described step (b), the estimated value that fractional part of frequency offset is the i time is current estimated value and estimation beforeThe mean value of value, that is:

Further, in described step (c), the variance of fractional part of frequency offset is drawn by following formula: $D_f\left(i\right)={|{\hat{f}}_F\left(i\right)-E_f\left(i\right)|}^2={|{\hat{f}}_F\left(i\right)-\underset{j=0}{\overset{N_E-1}{\Σ}}{\hat{f}}_F(i,j)|}^2,$ Its mean value is obtained by following formula: $V_f=\underset{j=0}{\overset{N_D}{\Σ}}{D}_f(i-j);$

As preferably, in described step (a) and step (b), the firstorder filter that is all 0.5 by parameter carries out average meterCalculate.

Compared with prior art, the present invention has following beneficial effect:

(1) times Carrier Synchronization Algorithm of the decimal in the present invention is accurately succinct, can effectively reduce decimal overtones band estimateResidual error, reduces error code, thereby makes system works in less residual error environment, and the stability that can improve system also can be furtherReduce the error rate of data, conceive very ingeniously, efficiently solve the deficiency of existing synchronized algorithm, there is outstanding substance spyPoint and marked improvement, be applicable to large-scale promotion application.

(2) on the basis that the present invention estimates at carrier wave frequency deviation single, also carried out average computing repeatedly, such calculatingMethod has further improved computational accuracy, has strengthened serviceability.

(3) to adopt parameter be that 0.5 firstorder filter carries out mean value computation in the present invention, greatly improved technical efficiency and standardExactness, the demand in realistic application.

Brief description of the drawings

Fig. 1 is schematic flow sheet of the present invention.

Fig. 2 is emulation schematic diagram one of the present invention.

Fig. 3 is emulation schematic diagram two of the present invention.

Detailed description of the invention

Below in conjunction with drawings and Examples, the invention will be further described, and embodiments of the present invention include but not limited toThe following example.

Embodiment

Carrier wave frequency deviation is not because the frequency of oscillation of transmitting terminal and receiving terminal is mated and produced, and it shows as in time domainA cumulative phase place shows as the overall offset of subcarrier on frequency domain. Because ofdm system need to ensure its subcarrierOrthogonality, thus very responsive to frequency shift (FS), higher to the requirement of carrier synchronization. In the present invention, we will receiveSignal be divided into carrier wave frequency deviation part and remaining part, therefore, establish that some OFDM of receiving after Complete Synchronization meetK time domain sampling point is: r (k)=s (k) exp (j2 π Δ fkT_s)+n (k)=s (k) exp (j2 π Δ fkT/N)+n (k), does oneA little conversion can draw: r (k)=s (k) exp (j2 π k (f_I+f_F)/N)+n (k), analyze from numeric field, because k is integer, because ofThis f_IWithout impact, and then can draw r (k)=s (k) exp (j2 π kf to signal time domain_F/N)+n(k)。

As can be seen from the above equation, obtain f_FNeed to eliminate time factor k, carry out time delay related operation. Due toHINOC system does not have pilot tone, so the algorithm based on pilot tone can not use, and targeting sequencing is not two sections of identical lettersNumber, thereby delay and correlation algorithm based on sequence can not use, and therefore estimates us for the fractional part of frequency offset of HiNOC systemCan only start with from cyclic prefix CP.

The algorithm of this part is generally to do based on being correlated with, and is applied in native system and carries out summation operation, and specific algorithm walksSuddenly be: $R_{\hat{n}}=\underset{k=1}{\overset{N_{CP}}{\Σ}}{r}^{*}(\hat{n}-k)r(\hat{n}-k+N),$ Wherein N is data segment, length 256, N_CPFor circulating prefix-length 16,For symbolThe OFDM symbol original position number synchronously obtaining; And then draw the single estimated value of fractional part of frequency offset?Here, accumulating operation is wanted to work as in phase place has been done on average.

According to emulation, in the time of Eb/No=20dB, the fractional part of frequency offset evaluated error of existing algorithm is at 0.01 gall nut intercarrierEvery in, although for low order system normally acceptable of this error, have following two reasons to make itAlso inapplicable in HINOC system.

First: HINOC system is high order system, its maximum modulation exponent number is 1024QAM, the fractional part of frequency offset meeting of 0.01 timesSystematic function is caused to larger impact, produce a large amount of error codes.

Second: in the frame structure of HINOC system, OFDM symbolic component does not contain pilot tone, therefore can only carry out one at frame headSecondary channel is estimated, and can not carry out channel estimating to each symbol, carries out channel equalization repeatedly, so can not when channel equalizationThe impact of correcting residual frequency deviation, makes residual frequency deviation accumulate the phase rotating bringing and constantly increases, until next frame head just can entangleJust, therefore, this cumulative phase rotating can cause adverse effect to systematic function.

In order to address the above problem, in the present invention we provide a kind of not only can reduce decimal overtones band estimateResidual error, improve the stability of a system, can also further reduce data error rate, allow same in less residual error environment of system worksStep algorithm.

As shown in Figure 1, the method that we adopt second order ring lock to judge is carried out the analysis of fractional part of frequency offset, mainly comprises followingStep:

(1) according to the single estimated value of fractional part of frequency offsetObtain its estimated value of the i time:Be the average of current estimated value and estimated value before; Wherein i is positive integer, and its value can be 1,2,3......N.

(2) repeated execution of steps (b), calculates the variance of gained fractional part of frequency offset, and calculates mean of variance; In the present inventionIn, the variance of fractional part of frequency offset is drawn by following formula: $D_f\left(i\right)={|{\hat{f}}_F\left(i\right)-E_f\left(i\right)|}^2={|{\hat{f}}_F\left(i\right)-\underset{j=0}{\overset{N_E-1}{\Σ}}{\hat{f}}_F(i,j)|}^2,$ ItsMean value is obtained by following formula: $V_f=\underset{j=0}{\overset{N_D}{\Σ}}{D}_f(i-j);$

(3) thresholding V is set, if V_f> V proceeds the single of fractional part of frequency offset and estimates, if V_f≤ V locks currentFractional part of frequency offset estimated value, and carry out compensate of frequency deviation with this estimated value all the time.

Can find out, on the basis that the present invention estimates at carrier wave frequency deviation single, carry out average computing repeatedly, carry with thisHigh estimated accuracy, the method for the averaging firstorder filter that namely parameter is 0.5, when the average of variance is less than a thresholdingTime, illustrate that the carrier wave frequency deviation estimating tends to be steady, at this moment lock, just can obtain a comparatively accurate fractional part of frequency offsetEstimate.

As shown in Figure 2 and Figure 3, we have carried out emulation to the algorithm in the present invention. Simulation parameter be Eb/No=10～25dB; Fractional part of frequency offset is 0.1, i.e. 6250Hz; Thresholding V=0.0005, multipath channel, ideal communication channel is estimated. Can be seen by figureGo out, with raising frequency deviation shortening locking time of signal to noise ratio, frequency deviation locking precision is totally improving, but has a fluctuation range.This is because emulation carries out is the locking of single, does not carry out a large amount of emulation and averages, so have certain randomProperty. As can also be seen from Figure, when signal to noise ratio is in the time that 17dB is above, residual frequency deviation is locked in 1Hz substantially, and when lockingBetween in 0.04 second. For the communication of this wired non sudden, such as cable television, 0.04 second delay be completely canWith what accept, even, under more high-precision requirement, can reduce again thresholding V, make residual frequency deviation be locked in lower interval,Its service behaviour is very good.

According to above-described embodiment, just can realize well the present invention.

Claims (3)

1. times carrier synchronization method of the decimal based on Cyclic Prefix, is characterized in that, comprises the following steps:

(a) signal receiving is divided into carrier wave frequency deviation part and remaining part, and carrier wave frequency deviation part is divided into little several timesFrequency deviation and integer frequency offset, and calculate the single estimated value of fractional part of frequency offset

(b) according to the single estimated value of fractional part of frequency offsetObtain its estimated value of the i time, wherein, i is positive integer;

(c) repeated execution of steps (b), calculates the variance of gained fractional part of frequency offset, and calculates mean of variance V_f；

(d) thresholding V is set, if current V_f> V continues execution step (a), if V_f≤ V locks current fractional part of frequency offset and estimatesEvaluation, and carry out compensate of frequency deviation with this estimated value all the time;

In described step (a), the single estimated value of decimal times carrier wave frequency deviationDrawn by following steps:

(a1) k the time domain sampling point r (k) that establishes the OFDM symbol receiving after Complete Synchronization is: r (k)=s (k) exp (j2 πkf_F/N)+n(k)；

(a2) summation of the related operation based on cyclic prefix CP drawsWherein N is data segmentLength 256, N_CPFor circulating prefix-length 16,The OFDM symbol original position obtaining for sign synchronization; And then draw little several timesThe single estimated value of frequency deviation

In described step (b), the estimated value that fractional part of frequency offset is the i time is current estimated value and the mean value of estimated value before, that is: ${\hat{f}}_F\left(i\right)=\left({\hat{f}}_F\right(i-1)+{\hat{f}}_F)/2.$

2. times carrier synchronization method of the decimal based on Cyclic Prefix according to claim 1, is characterized in that described step(c), in, the variance of fractional part of frequency offset is drawn by following formula: $D_f\left(i\right)=|{\hat{f}}_F\left(i\right)-E_f\left(i\right){|}^2=|{\hat{f}}_F\left(i\right)-\underset{j=0}{\overset{N_E-1}{\Σ}}{\hat{f}}_F(i-j){|}^2,$ Its mean value is obtained by following formula: $V_f=\underset{j=0}{\overset{N_D}{\Σ}}{D}_f(i-j).$

3. times carrier synchronization method of the decimal based on Cyclic Prefix according to claim 2, is characterized in that described step(a) and in step (b), the firstorder filter that is all 0.5 by parameter carries out mean value computation.