CN103298101B - A kind of code subcarrier synchronization realizing method of wide region - Google Patents

Abstract

The invention relates to a method for implementing wide-range code-assisted carrier synchronization, which divides the code-assisted carrier synchronization into two parts: coarse synchronization and fine synchronization. Firstly, based on the criterion of maximizing the evaluation function, the two-dimensional search window algorithm is used to limit the carrier offset within a certain range, and then the fine estimation algorithm with a larger parameter estimation range and lower complexity is selected to accurately realize carrier recovery. The advantages of the present invention are that it can greatly increase the synchronization range, has high estimation accuracy, can effectively realize carrier synchronization, and obtain decoding performance close to ideal synchronization.

Description

A Wide Range Code-Assisted Carrier Synchronization Method

technical field

The invention belongs to the technical field of demodulation and synchronization in communication, and in particular relates to a method for implementing wide-range code-assisted carrier synchronization.

Background technique

In the wireless communication system, the modulated data sequence is transmitted through various types of channels. Due to the influence of factors such as the instability of the transmitting and receiving oscillator frequency, oscillator drift, and mismatching of the transmitting and receiving filters, the received signal is inevitably introduced. carrier deviation. During coherent demodulation, the receiver must use various carrier synchronization methods to estimate the value of the carrier deviation as accurately and quickly as possible, and generate a coherent carrier with the same frequency and phase as the modulated carrier at the transmitting end, so as to restore the same frequency as the transmitting end. Modulates the data sequence. It can be seen that the carrier synchronization technology is a key technology in the receiver, and its performance is an important criterion for evaluating the performance of the receiver. How to effectively carry out carrier synchronization so as to demodulate the signal correctly and ensure the reliable transmission of information, solve the problem of This question is critical to the entire communication system.

On the one hand, in non-cooperative demodulation, the traditional Non-Data-Aided (NDA) carrier synchronization method cannot make full use of the implicit information in the signal, especially under the condition of low signal-to-noise ratio, the signal is affected by noise If the interference is large, the parameter estimation will have a large deviation. On the other hand, due to the low density parity check (LowDensityParityCheck, LDPC) code has a strong error correction ability in a low signal-to-noise ratio environment, and its performance is very close to the Shannon limit, it has been widely used in various communication systems. This makes Code-Aided (CA) carrier synchronization a research hotspot in recent years. It introduces decoding soft information into the estimation process of synchronization parameters, which can greatly reduce the SNR range in which the synchronization circuit can work effectively, making Communication lines that cannot be handled by conventional synchronization methods are demodulated with high quality. In addition, this code-assisted synchronization method can also be applied to cooperative demodulation, thereby further saving communication power.

In order to effectively achieve demodulation synchronization in a low SNR environment, it is usually necessary to use LDPC codes with large code length and low code rate. The code-assisted synchronization method can make full use of the coding information of a large number of symbols and has high estimation accuracy. However, it shows the problem of small parameter estimation range. This is because the large frequency offset and phase offset will greatly reduce the signal power input to the decoder, resulting in unreliable decoding of soft information, which in turn makes the synchronizer and decoder Neither of them can converge, especially the large residual frequency offset will lead to serious deterioration of LDPC decoding performance. Secondly, the code-assisted synchronization method needs to use the soft information output by the LDPC decoder to iteratively update the estimated value of the synchronization parameters, and the LDPC code usually uses the logarithmic domain Belief Propagation (BP) principle to realize iterative decoding, and the frequency estimation usually uses The search for the maximum value of , introduces a high amount of calculation, and under the condition of low signal-to-noise ratio, more iterations are needed to make the decoder and synchronizer converge.

Contents of the invention

technical problem to be solved

In order to avoid the deficiencies of the prior art, the present invention proposes a wide-range code-assisted carrier synchronization implementation method, which is a code-assisted synchronization scheme that can take into account the carrier synchronization parameter estimation range, estimation accuracy, and complexity requirements, so that the code The advantage of the auxiliary synchronization method in the environment of low signal-to-noise ratio can be brought into full play, and it is suitable for the carrier synchronization of the communication system under the condition of low signal-to-noise ratio.

Technical solutions

A wide range of code-assisted carrier synchronization implementation method is characterized in that the steps are as follows:

Step 1: Take the maximum phase and frequency values that can be shifted by the received signal sequence r _k as the boundary of the phase-frequency two-dimensional search space, and select n discrete phase-frequency points (θ _i , Δf _j ), and use the following formula to compensate the carrier deviation of the received signal:

${{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&theta;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;k\&Delta;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}$

In the formula, k=0,1,2...K-1, K is the symbol length of channel coding, and T is the symbol interval period;

Demodulate and decode n compensated signals r _k ′ sequentially;

Step 2: Take the absolute value of the K pieces of soft information L(c _k ) output by each group of decoding and then make a threshold judgment, select |L(c _k )| that is greater than the threshold value to accumulate and average to obtain Then compare n The size of the value, choose The phase-frequency point corresponding to the maximum is used as the estimated value of coarse synchronization And its corresponding compensated signal r _k ′ is the input signal y _k of carrier fine synchronization;

Step 3: Initialize the fine synchronization of the carrier, set the number of iterations i=0, and the counter Let the maximum number of iterations of fine synchronization be 1;

Step 4: Leverage and Update the input signal vector for carrier fine synchronization:

${\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; (</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&theta;</span>}}}_{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;k\&Delta;</span>}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}$

on signal Demodulate and decode to get the output soft information L _i (c _k ), and then use the same method as step 2 to get

Step 5: When i>1, calculate the i-th iteration and the i-1th iteration difference, if Add 1 to the counter count, otherwise the counter is cleared; if i=1, skip this step directly and go directly to step 6;

Step 6: Use the L _i (c _k ) corresponding to |L _i (c _k )| greater than the threshold value in step 4 to calculate the mean value of the posterior probability of all possible symbols on the modulation constellation diagram by the following formula

In the formula, M is the number of constellation points, a _k is the received modulation symbol sequence, A _m is the value of the mth point on the modulation constellation diagram, and tanh is the hyperbolic tangent operation;

Step 7: Leverage For frequency offset estimation, use the simplified EM estimation method

The estimated value of the frequency offset can be obtained by searching the maximum value. The search interval uses the constant convergence characteristic of the EM algorithm to adaptively reduce the size of the next frequency search interval according to the results of the first two iterations. The search interval for the parameters is:

$\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&Delta;</span>}\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span>$

in, $f_x^{(i+1)}=|\mathrm{\&Delta;}{\hat{f}}_x^{(i+1)}-\mathrm{\&Delta;}{\hat{f}}_x^{\left(i\right)}|;$

Step 8: Leverage To make a bias estimate, And let n=n+1;

Step 9: Repeat the above steps (4) to (8) until count=P or i=I, and obtain the precise estimated value of the carrier deviation and Exit the iterative loop and execute the next step;

Step 10: Use the final fine estimate of carrier offset and Compensate the fine synchronous input signal y _k to get z _k : After demodulation, decoding and judgment, the original information of the sending end can be recovered.

Using the frequency offset estimation method based on the autocorrelation function to replace the simplified EM estimation method in step 7, first by The modulation information in the signal is removed then ask The autocorrelation function value R(m), and then use a rectangular window with a length of K/2, according to the smoothing of the differential complex values of K/2-1 adjacent autocorrelation functions and the order of taking phase angles, the corresponding The frequency offset estimation formulas are: $\begin{array}{c}\mathrm{\&Delta;}{\hat{f}}_x^{\left(i\right)}=\frac{1}{2\mathrm{\&pi;T}}\arg \left\{\underset{m=0}{\overset{K/2}{\mathrm{\&Sigma;}}}R\left(m\right)R^{*}(m-1)\right\}\\ \mathrm{\&Delta;}{\hat{f}}_x^{(i+1)}=\frac{1}{2\mathrm{\&pi;T}}\underset{m=2}{\overset{K/2}{\mathrm{\&Sigma;}}}\arg \left[R\left(m\right)R^{*}(m-1)\right]\end{array}.$

The threshold of the step is L _m , and L _m =2-4.

The said λ=0.01～0.1.

Beneficial effect

The invention proposes a method for implementing wide-range code-assisted carrier synchronization, which divides the code-assisted carrier synchronization into two parts: coarse synchronization and fine synchronization. Firstly, based on the criterion of maximizing the evaluation function, the two-dimensional search window algorithm is used to limit the carrier offset within a certain range, and then the fine estimation algorithm with a larger parameter estimation range and lower complexity is selected to accurately realize carrier recovery. The advantages of the present invention are that it can greatly increase the synchronization range, has high estimation accuracy, can effectively realize carrier synchronization, and obtain decoding performance close to ideal synchronization. The features of the present invention are:

(1) In the carrier coarse synchronization part, the absolute value of the decoding output soft information L(c _k ) is used as a measure of the reliability of the bit decision, and the average value is an important criterion for judging whether the synchronizer and the decoder tend to converge, namely When the carrier offset estimated and used for compensation is most accurate, the mean value of the absolute value of L(c _k ) to reach maximum.

⑵Shorten the synchronization delay introduced by the search in coarse synchronization through parallel processing.

(3) By setting the threshold value to select symbols with high decision reliability, and by selectively performing decoding decision feedback to reduce the number of unreliable calculation points and further reduce the number of multiplication operations. This feature is reflected in both step 2 and step 4.

(4) Propose a new judgment criterion for stopping the iteration early, that is, according to Whether the value of the value tends to be stable determines whether the iteration is performed or not. Compared with other fixed number of iterations, unnecessary iterations are avoided and the amount of calculation is reduced.

(5) In view of the high calculation amount introduced in the search process in the fine frequency offset estimation, two solutions are given. One is to replace the fixed window with an adaptively reduced search window. The gradually reduced search interval will inevitably lead to Smaller calculation amount; the second is to use the frequency offset estimation method of the autocorrelation function, using the posterior mean After demodulation, the frequency offset estimation is transformed into the frequency estimation problem of complex single-frequency signal in complex Gaussian white noise, which effectively avoids the large computational complexity caused by iterative search.

Description of drawings

Fig. 1 is the realization block diagram of wide-ranging code-assisted carrier synchronization;

Fig. 2 is the realization block diagram of carrier coarse synchronization;

Fig. 3 is the realization block diagram of carrier fine synchronization;

Fig. 4 is a realization block diagram of frequency offset estimation by autocorrelation function method in carrier fine synchronization.

detailed description

Now in conjunction with embodiment, accompanying drawing, the present invention will be further described:

The code-assisted carrier synchronization realized by the present invention is composed of two modules, one is coarse synchronization based on the criterion of maximizing the evaluation function, and the other is fine synchronization using fine estimation algorithm with large parameter estimation range and low complexity.

Considering the BPSK modulation system under the LDPC coding method, the signal passes through the additive white Gaussian noise channel, under the assumption of ideal symbol timing recovery, ideal frame synchronization, and neglecting the signal gain and intersymbol interference, only the carrier offset of the receiving end is studied. influences.

Figure 1 is a block diagram of the realization of the principle of this example. The soft information L(c _k ) fed back by the decoder is used to carry out coarse synchronization and fine synchronization of the carrier in turn, and the received signal is compensated by the large frequency offset and large phase offset obtained by rough estimation. , to limit the remaining carrier offset of the received signal within a certain range, then use the fine estimation algorithm to estimate and compensate it, and then perform BPSK demodulation, and then restore the original sent information after decoding and hard decision.

Figure 2 is a block diagram of the realization of the principle of carrier coarse synchronization in this example. In the two-dimensional phase-frequency space determined by the maximum carrier offset, the search window algorithm is used, that is, n uniformly discrete phase-frequency points (θ _i , Δf _j ), respectively use them to compensate the carrier offset of the received signal, after demodulation and decoding, for each group of decoding output soft information L(c _k ), set the threshold L _m , when |L( When c _k )|≤L _m , directly set L(c _k ) to zero, and for the remaining symbols with high decision reliability, decode the mean value of the absolute value of soft information As a new evaluation function, choose The phase-frequency point corresponding to the maximum value is used as a rough estimate of the carrier parameter The corresponding compensated signal is the carrier finely synchronized input signal y _k .

Fig. 3 is a block diagram of the implementation principle of carrier fine synchronization in this example, which mainly consists of four parts. One is to selectively perform decoding decision feedback, which corresponds to the figure in which a threshold L _m decision device is added after the LDPC decoder. When |L(c _k )|≤L _m , L( c _k ) is set to zero. The second is to use the new synchronous iteration to stop the decision device to reduce unnecessary iterations. For the symbols with high reliability of the remaining decisions in the previous step, take the mean value of the absolute value of their decoding soft information By setting two detection An important parameter whether it tends to be stable, one is the threshold λ, which is used to characterize Whether the change is small enough; the other is the depth factor P, which is used to judge Whether the gradually plateauing trend is continuous. The specific judgment method is to set a small reference value λ in advance, and after the i-th iteration, take this iteration and the previous P iterations and every two adjacent iterations value subtraction to take the absolute value, if all the results are less than λ, we consider has plateaued, thus stopping synchronous iterations. The third is to provide two simpler estimators for the higher calculation amount introduced by the search process in the original frequency offset estimator. The first is the simplification based on the original EM frequency offset estimator, that is, the fixed window is replaced by an adaptively reduced search window in the estimation process. Specifically, the size of the next frequency search interval is adaptively adjusted according to the results of the previous two iterations, that is, $\mathrm{\&Delta;}{\hat{f}}^{(\mathrm{no}+2)}\&Element;({\mathrm{\&Delta;}\hat{f}}^{(\mathrm{no}+1)}-f_c^{(\mathrm{no}+1)},\mathrm{\&Delta;}{\hat{f}}^{(\mathrm{no}+1)}+f_c^{(\mathrm{no}+1)}),$ in, $f_c^{(\mathrm{no}+1)}=|\mathrm{\&Delta;}{\hat{f}}^{(\mathrm{no}+1)}-\mathrm{\&Delta;}{\hat{f}}^{\left(\mathrm{no}\right)}|.$ The second is a frequency offset estimator based on an autocorrelation function, see Figure 4 for details. The fourth is a phase bias estimator based on the maximum likelihood criterion. Estimate the accurate carrier offset by frequency offset and phase offset and Afterwards, the received signal y _k is compensated to realize complete carrier synchronization.

Figure 4 is a block diagram of the principle implementation of the frequency offset estimator based on the autocorrelation function in the fine synchronization of the carrier in this example. First, the posterior mean value is used Multiply with the input signal y _k to remove the modulation information in the signal to get x _k , and then calculate the autocorrelation function value R(m) of x _k , and conjugate all successively adjacent autocorrelation function values within a certain window After multiplication, the phase angle is taken to estimate the frequency difference, and these mutually statistically independent estimated values can be weighted to obtain an estimation algorithm with good anti-noise performance. For simplicity, this example uses a rectangular window with a length of K/2. According to The order of smoothing the differential complex values of K/2-1 adjacent autocorrelation functions and taking the phase angle is different, and the frequency offset estimation formula is divided into the following two types:

$\begin{array}{c}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;T</span>}}\mathrm{<span\; class="notranslate">\; arg</span>}<span\; class="notranslate">\; \{</span>\underset{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; R</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>\end{array}$

$\begin{array}{c}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}}_{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;T</span>}}\underset{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\overset{\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; arg</span>}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; R</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>\end{array}$

All of them can effectively realize the fast estimation of large frequency offset under the condition of low signal-to-noise ratio with relatively low complexity. Through loop iterations, the accurate value of frequency offset estimation is finally obtained

Claims (4)

1. a wide range of code-assisted carrier synchronization method, characterized in that the steps are as follows: Step 1: Take the maximum phase and frequency values that can be shifted by the received signal sequence r _k as the boundary of the phase-frequency two-dimensional search space, and select n discrete phase-frequency points (θ _i , Δf _j ), and use the following formula to compensate the carrier deviation of the received signal: In the formula, k=0,1,2...K-1, K is the symbol length of channel coding, and T is the symbol interval period; Demodulate and decode n compensated signals r _k ′ sequentially; Step 2: Take the absolute value of the K pieces of soft information L(c _k ) output by each group of decoding and then make a threshold judgment, select |L(c _k )| that is greater than the threshold value to accumulate and average to obtain Then compare n The size of the value, choose The phase-frequency point corresponding to the maximum is used as the estimated value of coarse synchronization And its corresponding compensated signal r _k ′ is the input signal y _k of carrier fine synchronization; Step 3: Carrier fine synchronization is initialized, the number of iterations i=0, the counter count=0, Let the maximum number of iterations of fine synchronization be 1; Step 4: Leverage and Update the input signal vector for carrier fine synchronization: on signal Demodulate and decode to get the output soft information L _i (c _k ), and then use the same method as step 2 to get Step 5: When i>1, calculate the i-th iteration and the i-1th iteration difference, if Add 1 to the counter count, otherwise the counter is cleared; if i=1, skip this step directly and go directly to step 6; Step 6: Use the L _i (c _k ) corresponding to |L _i (c _k )| greater than the threshold value in step 4 to calculate the mean value of the posterior probability of all possible symbols on the modulation constellation diagram by the following formula In the formula, M is the number of constellation points, a _k is the received modulation symbol sequence, A _m is the value of the mth point on the modulation constellation diagram, and tanh is the hyperbolic tangent operation; Step 7: Leverage For frequency offset estimation, use the simplified EM estimation method The estimated value of the frequency offset can be obtained by searching for the maximum value. The search interval uses the constant convergence characteristic of the EM algorithm to adaptively reduce the size of the next frequency search interval according to the results of the first two iterations. The search interval for the parameters is: in, Step 8: Leverage To make a bias estimate, And let n=n+1; Step 9: Repeat above-mentioned steps (4)～(8), until count=P or i=1 time, obtain carrier deviation accurate estimated value and Exit the iterative loop and execute the next step; Step 10: Use the final fine estimate of carrier offset and Compensate the fine synchronous input signal y _k to get z _k : After demodulation, decoding and judgment, the original information of the sending end can be recovered. 2. according to the wide-ranging code-assisted carrier synchronization realization method of claim 1, it is characterized in that: adopt the simplified EM estimation method of step 7 to replace the frequency offset estimation method based on autocorrelation function, at first by The modulation information in the signal is removed then ask The autocorrelation function value R(m), and then use a rectangular window with a length of K/2, according to the smoothing of the differential complex values of K/2-1 adjacent autocorrelation functions and the order of taking phase angles, the corresponding The frequency offset estimation formulas are: 3. The wide range code-assisted carrier synchronization implementation method according to claim 1, characterized in that: the threshold in step 4 is L _m , and L _m =2-4. 4. The wide range code-assisted carrier synchronization implementation method according to claim 1, characterized in that: said λ=0.01˜0.1.