CN107276956A - Carrier synchronization method and device - Google Patents

Abstract

The present invention provides a carrier synchronization method, comprising: performing a cross-correlation operation on two training codes extracted from a received signal, wherein the received signal is obtained from an initial frequency domain The time-domain data signal generated by the inverse Fourier transform of the data signal is composed of the time-domain training sequence signal and the time-domain training sequence signal includes the training sequence and data based on the training code, wherein the training sequence bandwidth is greater than the data bandwidth and training The power spectral density of the sequence is lower than the power spectral density of the data; and frequency offset correction is performed on the received signal based on the frequency offset. The invention improves the frequency offset precision of the system and reduces the bit error rate of the whole system.

Description

Carrier synchronization method and device

technical field

The present invention generally relates to a wireless communication system, and in particular to a carrier synchronization method and device.

Background technique

Wireless communication networks are widely deployed to provide various communication services such as voice, video, packet data, messaging, broadcast, and so on. These wireless networks may be multiple-access networks capable of supporting multiple users by sharing available network resources. Examples of such multiple-access networks include Code Division Multiple Access (CDMA) networks, Time Division Multiple Access (TDMA) networks, Frequency Division Multiple Access (FDMA) networks, Orthogonal FDMA (OFDMA) networks, and Single-Carrier FDMA (SC-FDMA) networks. )network.

With the ever-increasing demand for global mobile communications, the frequency resources of wireless communications are becoming increasingly tight. Therefore, in addition to the above-mentioned traditional wireless communication systems with high spectrum efficiency based on TDM (Time Division Multiplexing) and FDM (Frequency Division Multiplexing), more aggressive communication schemes with higher frequency spectrum utilization have also been proposed.

An overlapped time division multiplexing (Overlapped Time Division Multiplexing, OvTDM) system is just such a scheme for improving the spectrum efficiency of the system. In the OvTDM system, symbols do not need to be isolated from each other, but can have a strong overlapping. In other words, the OvTDM system artificially introduces overlap between symbols, and uses multiple symbols to transmit data sequences in parallel in the time domain, greatly improving spectrum utilization.

An overlapping frequency division multiplexing (Overlapped Frequency Division Multiplexing, OvFDM) system is another solution for improving the spectral efficiency of the system. In an OvFDM system, there can be stronger overlap between subcarrier frequency bands than in OFDM. Based on the OFDM system, the utilization rate of the frequency spectrum is further improved through a higher degree of overlapping between sub-frequency bands in the frequency domain.

Although the above-mentioned OvTDM system and OvFDM system have corresponding reception demodulation schemes to eliminate the interference caused by signal overlap in the time domain or frequency domain, the significant increase in spectrum utilization still puts forward higher requirements for signal reception.

Therefore, the OvTDM system and the OvFDM system need a higher-performance network access solution. However, the m-sequence carrier synchronization method adopted by the existing communication system cannot meet the requirements.

Contents of the invention

A brief summary of one or more aspects is presented below to provide a basic understanding of these aspects. This summary is not an exhaustive overview of all contemplated aspects and is intended to neither identify key or critical elements of all aspects nor attempt to delineate the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description that is presented later.

The purpose of the present invention is to provide a carrier synchronization method for the defects that the carrier frequency deviation is large and the synchronization result is inaccurate due to its poor autocorrelation and cross-correlation characteristics when the existing communication system uses m-sequence for carrier synchronization. Methods and devices to solve the above problems.

According to an aspect of the present invention, a carrier synchronization method is provided, including:

Perform a cross-correlation operation on the two training codes extracted from the received signal to obtain the frequency offset of the carrier between the receiving end and the transmitting end, where the received signal is the time domain data generated by the initial frequency domain data signal after Fourier inverse transform The signal and the training sequence signal in the time domain and the training sequence signal in the time domain include the training sequence and data based on the training code, wherein the training sequence bandwidth is greater than the data bandwidth and the power spectral density of the training sequence is lower than the power spectral density of the data ;as well as

Frequency offset correction is performed on the received signal based on the frequency offset.

In an example, the training code includes m-sequence, Golomb code, CAN code, or LAS code.

In one example, the training sequence bandwidth is 5 times, 10 times, 15 times or more than the data bandwidth.

In one example, the training sequence includes two LAS short codes [X _las ] _SN , where SN is the length of the LAS short code, wherein the cross-correlation operation is performed on the two training codes extracted from the received signal to obtain The frequency offset of the carrier between the transmitter and the transmitter includes:

performing a cross-correlation operation on the two LAS short codes extracted from the received signal to obtain a coarse frequency offset of the carrier between the receiving end and the transmitting end, and

Performing frequency offset correction on the received signal based on the frequency offset includes:

A primary frequency offset correction is performed on the received signal based on the coarse frequency offset.

In an example, the training sequence includes [0] _SN , [X _las ] _SN , [0] _SN , [X _las ] _SN , where [0] _SN is a 0 sequence with a length of SN.

In one example, the training sequence further includes two LAS long codes [X _las ] _LN , where LN is the length of the LAS long code, wherein the cross-correlation operation is performed on the two training codes extracted from the received signal to obtain the received The frequency offset of the carrier between the end and the transmitter also includes:

performing a cross-correlation operation on two LAS long codes extracted from the received signal subjected to this initial frequency offset correction to obtain a fine frequency offset of the carrier between the receiver and the transmitter, and

Performing frequency offset correction on the received signal based on the frequency offset further includes:

Based on the coarse frequency offset and the fine frequency offset, a second frequency offset correction is performed on the received signal after the primary frequency offset correction.

In an example, performing secondary frequency offset correction on the received signal after the primary frequency offset correction based on the coarse frequency offset and the fine frequency offset includes:

The secondary frequency offset correction is performed on the received signal after the primary frequency offset correction based on the sum of the coarse frequency offset and the fine frequency offset.

In one example, the training sequence includes [0] _SN , [X _las ] _SN , [0] _SN , [X _las ] _SN , [X _las ] _LN , [X _las ] _LN , where [0] _SN is the length is the 0 sequence of SN.

In an example, the two LAS short codes are also used for timing synchronization, and the extraction positions of the two LAS short codes and the two LAS long codes are determined based on timing synchronization results.

According to another aspect of the present invention, a carrier synchronization device is provided, including:

a cross-correlation calculation unit, configured to perform a cross-correlation operation on two training codes extracted from a received signal, wherein the received signal is obtained by Fourier inverse The time-domain data signal generated after transformation is composed of the time-domain training sequence signal and the time-domain training sequence signal includes the training sequence and data based on the training code, wherein the training sequence bandwidth is greater than the data bandwidth and the power spectral density of the training sequence below the power spectral density of the data; and

A frequency correction unit, configured to perform frequency offset correction on the received signal based on the frequency offset.

In an example, the training code includes m-sequence, Golomb code, CAN code, or LAS code.

In one example, the training sequence bandwidth is 5 times, 10 times, 15 times or more than the data bandwidth.

In an example, the training sequence includes two LAS short codes [X _las ] _SN , where SN is the length of the LAS short code, wherein, the correlation calculation unit is further used to compare the two LAS short codes extracted from the received signal The code performs a cross-correlation operation to obtain a coarse frequency offset of the carrier between the receiver and transmitter, and

The frequency correction unit is further configured to perform primary frequency offset correction on the received signal based on the rough frequency offset.

In an example, the training sequence includes [0] _SN , [X _las ] _SN , [0] _SN , [X _las ] _SN , where [0] _SN is a 0 sequence with a length of SN.

In an example, the training sequence further includes two LAS long codes [X _las ] _LN , where LN is the length of the LAS long code, wherein the correlation calculation unit is further used to extract from the received The two LAS long codes of the signal perform a cross-correlation operation to obtain a fine frequency offset of the carrier between the receiver and transmitter, and

The frequency correction unit is further configured to perform secondary frequency offset correction on the received signal after the primary frequency offset correction based on the coarse frequency offset and the fine frequency offset.

In an example, the frequency correction unit is further configured to perform the secondary frequency offset correction on the received signal after the primary frequency offset correction based on the sum of the coarse frequency offset and the fine frequency offset.

In one example, the training sequence includes [0] _SN , [X _las ] _SN , [0] _SN , [X _las ] _SN , [X _las ] _LN , [X _las ] _LN , where [0] _SN is the length It is the 0 sequence of SN.

In an example, the two LAS short codes are also used for timing synchronization, and the extraction positions of the two LAS short codes and the two LAS long codes are determined based on timing synchronization results.

The present invention has the following beneficial effects: the present invention designs the LAS code training sequence in the communication system, utilizes the characteristics that the LAS code autocorrelation function is an ideal impact function at the origin, is zero everywhere outside the origin, and the cross-correlation function is zero everywhere , use the LAS code for carrier synchronization in signal reception processing. When the LAS code is used for carrier synchronization, the short LAS code coarse synchronization and the long LAS code fine synchronization can improve the frequency offset accuracy of the system, lay a foundation for the subsequent channel estimation process and decoding process, and reduce the error of the entire system. code rate.

Description of drawings

The above-mentioned features and advantages of the present invention can be better understood after reading the detailed description of the embodiments of the present disclosure in conjunction with the following drawings. In the drawings, components are not necessarily drawn to scale, and components with similar related properties or characteristics may have the same or similar reference numerals.

Fig. 1 shows the block diagram of the transmitting end modulation module of OvTDM system;

Fig. 2 shows the block diagram of the signal preprocessing module of the receiving end of OvTDM system;

Fig. 3 shows the block diagram of the receiver sequence detection module of OvTDM system;

Fig. 4 shows the modulation module block diagram of the transmitting end of OvFDM system;

Fig. 5 shows the block diagram of the signal preprocessing module of the receiving end of OvFDM system;

Fig. 6 shows the block diagram of the signal detection module of the receiving end of OvFDM system;

Fig. 7 shows the autocorrelation characteristic of M sequence;

Fig. 8 shows the autocorrelation characteristic of LAS code;

Figure 9 shows a distribution diagram of autocorrelation results for timing synchronization;

Fig. 10 shows a schematic diagram of a training sequence in which two peaks are detected;

FIG. 11 shows a block diagram of a timing synchronization unit at a receiving end according to an aspect of the present invention;

Fig. 12 shows a flowchart of a timing synchronization method according to an aspect of the present invention;

FIG. 13 shows a block diagram of a carrier synchronization unit according to an aspect of the present invention;

FIG. 14 shows a flowchart of a carrier synchronization method according to an aspect of the present invention;

FIG. 15 shows a flowchart of a carrier synchronization method according to an aspect of the present invention;

Figure 16 shows a schematic diagram of the arrangement of multipath channels;

Fig. 17 shows the bandwidth and power spectral density relationship graph of training sequence and data according to an aspect of the present invention; And

Fig. 18 shows a schematic diagram of frequency spectrum when two carrier signals transmit data simultaneously according to an aspect of the present invention.

detailed description

The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments. Note that the aspects described below in conjunction with the drawings and specific embodiments are only exemplary, and should not be construed as limiting the protection scope of the present invention.

In addition to being used in OvTDM and OvFDM systems, the technologies described in this paper can also be widely used in practical mobile communication systems, such as TD-LTE, TD-SCDMA and other systems, and can also be widely used in satellite communication, microwave line-of-sight communication In any wireless communication system such as , scattering communication, atmospheric optical communication, infrared communication and aquatic communication. The terms "network" and "system" are often used interchangeably.

The continuous development of mobile communications and the emergence of new services have put forward higher and higher requirements for data transmission rates, but the frequency resources of mobile communications are very limited. How to use limited frequency resources to achieve high-speed data transmission has become a major challenge for today's mobile communications technology. an important question of

The above-mentioned OvTDM and OvFDM systems are exactly such solutions that can greatly improve spectrum utilization. The following briefly introduces the sending and receiving process of the OvTDM system.

The OvTDM system utilizes multiple symbols to transmit data sequences in parallel in the time domain. At the transmitting end, a transmission signal with multiple symbols overlapping each other in the time domain is formed, and at the receiving end, according to the one-to-one correspondence between the transmission data sequence and the time waveform of the transmission data sequence, the received signal is detected according to the data sequence in the time domain. The OvTDM system actively utilizes these overlaps to generate coding constraints, thus greatly improving the system's spectral efficiency.

Fig. 1 shows a block diagram of a modulation module at a transmitting end of an OvTDM system. The modulation module 100 at the transmitting end may include a digital waveform generation unit 110 , a shift register unit 120 , a multiplication unit 130 and an addition unit 140 .

First, the digital waveform generating unit 110 digitally designs and generates the first modulated signal envelope waveform h(t) of the transmitted signal, and the shift register unit 120 shifts the envelope waveform h(t) by a specific time to form For the envelope waveform h(ti×ΔT) of the modulated signal at other times, the multiplication unit 130 multiplies the parallel symbol x _i to be transmitted with the envelope waveform h(ti×ΔT) at the corresponding time to obtain the modulated signal at each time The to-be-sent signal waveform x _i h(ti×ΔT). The adding unit 140 superimposes the formed waveforms to be transmitted to form a transmission signal waveform.

The receiving end of the OvTDM system is mainly divided into a signal preprocessing module 200 and a sequence detection module 300 . FIG. 2 shows a block diagram of a signal preprocessing module 200 at the receiving end of the OvTDM system. The signal preprocessing module is used to assist in forming a synchronously received digital signal sequence in each frame. As shown in the figure, the signal preprocessing module may include a synchronization unit 210 , a channel estimation unit 220 , and a digitization processing unit 230 .

The synchronization unit 210 is used to form symbol synchronization for the received signal in the time domain to maintain synchronization with the system, mainly including timing synchronization and carrier synchronization. After the synchronization is completed, the channel estimation unit 220 performs channel estimation on the received signal, so as to estimate the parameters of the actual transmission channel. The digital processing unit 230 is used for digital processing the received signal in each frame, so as to form a received digital signal sequence suitable for sequence detection by the sequence detection part.

After the preprocessing, sequence detection can be performed on the received signal in the sequence detection module 300, the received waveform is cut according to the waveform transmission time interval, and the cut waveform is decoded according to a certain decoding algorithm. Fig. 3 shows a block diagram of a receiver sequence detection module of an OvTDM system. As shown in the figure, the sequence detection module 300 may include an analysis storage unit 310 , a comparison unit 320 , and a reserved path storage unit and a Euclidean distance storage unit 330 . During the detection process, the analysis and storage unit makes the complex convolution coding model and the trellis diagram of the OvTDM system, lists all the states of the OvTDM system, and stores them. The comparison unit searches out the path with the minimum Euclidean distance to the received digital signal according to the trellis diagram in the analysis storage unit, and the reserved path storage unit and the Euclidean distance storage unit are used to store the reserved path and the Euclidean distance outputted by the comparison unit respectively. distance or weighted Euclidean distance. One reserved path storage unit and one Euclidean distance storage unit need to be prepared for each stable state. The length of the reserved path storage unit may preferably be 4K˜5K. The Euclidean distance storage unit preferably only stores relative distances.

Fig. 4 shows a block diagram of a modulation module at a transmitting end of an OvFDM system. The OvFDM modulation module at the transmitting end may include a modulated carrier spectrum generation unit 410 , a carrier spectrum shift unit 420 , a multiplication unit 430 , an addition unit 440 , and an inverse Fourier transform unit 450 .

First, the modulated carrier spectrum generating unit 410 is designed to generate an envelope spectrum signal H(f) of a subcarrier, and the carrier spectrum shift unit 420 shifts the envelope spectrum signal H(f) by a specific carrier spectrum interval ΔB sequentially, to obtain The envelope spectrum signal of the next subcarrier is obtained, and the frequency shift of the envelope spectrum signal of the next subcarrier is ΔB, and the spectrum waveform H(f-i×ΔB) of all subcarriers with a spectrum interval of ΔB is obtained sequentially.

The multiplication unit 430 multiplies the multi-channel parallel symbols X _i to be transmitted by the generated corresponding sub-carrier spectrum waveforms H (fi × ΔB) to obtain multiple modulation signal spectrums X _i H( fi × ΔB).

The addition unit 440 superimposes the formed multiple modulation signal spectrum to form the spectrum of the complex modulation signal Finally, the inverse Fourier transform unit 450 performs an inverse discrete Fourier transform on the frequency spectrum of the generated complex modulation signal to finally form a complex modulation signal Signal(t) _TX =ifft(S(f)) in the time domain.

The receiving end of the OvFDM system is mainly divided into a signal preprocessing module 500 and a signal detection module 600 . Fig. 5 shows a block diagram of a signal preprocessing module at the receiving end of the OvFDM system. As shown in the figure, the preprocessing module may include a synchronization unit 510 , a channel estimation unit 520 , and a digitization processing unit 530 .

The synchronization unit 510 is used to form symbol synchronization for the received signal in the time domain to maintain synchronization with the system, mainly including timing synchronization and carrier synchronization. After the synchronization is completed, the channel estimation unit 520 performs channel estimation on the received signal, so as to estimate the parameters of the actual transmission channel. The digitization processing unit 530 is used for sampling and quantizing the received signal in each symbol time interval, so that it becomes a digital signal sequence.

After preprocessing, the received signal can be detected in the signal detection module 600 . Fig. 6 shows a block diagram of a signal detection module 600 at the receiving end of the OvFDM system. As shown in the figure, the signal detection module 600 may include a Fourier transform unit 610 , a frequency segmentation unit 620 , a convolutional encoding unit 630 , and a data detection unit 640 . The Fourier transform unit 610 is used to convert the preprocessed time domain signal into a frequency domain signal, that is, perform Fourier transform on the received digital signal sequence of each time symbol interval to form the actual received signal spectrum of each time symbol interval. The frequency segmentation unit 620 is configured to segment the actual received signal spectrum of each time symbol interval in the frequency domain at a frequency interval ΔB to form the actual received signal segmented spectrum. The convolutional encoding unit 630 is used to form a one-to-one correspondence between the frequency spectrum of the received signal and the transmitted data symbol sequence. The data detection unit 640 is configured to detect the data symbol sequence according to the one-to-one correspondence formed by the convolutional coding unit.

The processing procedures of the sending and receiving ends of the OvTDM system and the OvFDM system have been introduced above. Although the above-mentioned OvTDM system and OvFDM system have corresponding reception demodulation schemes to eliminate the interference caused by signal overlap in the time domain or frequency domain, the significant increase in spectrum utilization still puts forward higher requirements for signal reception.

A training sequence needs to be designed in a general communication system, and its main function is to process the signal after it is received, so as to realize timing synchronization, carrier synchronization and channel estimation at the same time. Timing synchronization, carrier synchronization and channel estimation are the three most important links for correct reception at the receiving end. Therefore, the design of training symbols is very important, especially for communication systems with ultra-high spectrum efficiency such as OvTDM and OvFDM systems. If any of these three steps has a large error, it will have a great impact on the entire system, and the subsequent decoding process will be meaningless.

At present, communication systems often use M sequences as training sequences. Due to the poor autocorrelation and cross-correlation characteristics of M sequences, the success rate of the system synchronization process is low and the network access is slow. Figure 7 shows the autocorrelation characteristics of the M sequence. It can be seen from the figure that pulses will appear at intervals of a certain period of time in the autocorrelation characteristics, and its autocorrelation characteristics are not very good. Therefore, in the signal processing process, the synchronization accuracy of time and frequency is poor, which reduces the success rate and access speed of users accessing the network, and deteriorates user experience.

According to an aspect of the present invention, LAS codes are used to design training sequences in OvTDM system and OvFDM system. After research, it is found that the LAS code has the characteristics that the autocorrelation function is an ideal impulse function at the origin, is zero everywhere outside the origin, and the cross-correlation function is zero everywhere. This is an extremely favorable property for training sequences.

The LAS (Large Area Synchronized) code is composed of a series of pulses and 0-value pulse intervals of different lengths, which can be expressed as (N, K, L), where N represents the number of pulses, and K represents the interval between pulses The shortest interval length, L represents the code length. The pulse is generated by a complete complementary orthogonal code, and its characteristic is that the autocorrelation function is an ideal impulse function at the origin, and it is zero everywhere outside the origin, while the cross-correlation function is zero everywhere. This feature of LAS code is used in OvTDM system and OvFDM system, which can improve the synchronization success rate and access speed of the whole system.

The following briefly introduces the generation method of LAS code.

The complete complementary orthogonal code has a dual relationship, and the generation method is to solve another pair of shortest basic complementary codes that are completely orthogonal and complementary to it based on the shortest basic complementary code. In this case, the basic short code +++- is used to generate a complete complementary orthogonal code, and the generation process is as follows:

C ₀ =[1 1], corresponding to ++, S ₀ =[1 -1], corresponding to +-, according to C ₀ and S ₀ , calculate their complementary codes C ₁ and S ₁ respectively. C ₁ is obtained by inverting S ₀ , and S ₁ is obtained by inverting C ₀ and negating it. The code in matlab is expressed as:

C ₁ =fliplr(S ₀ ), S ₁ =−1*conj(fliplr(C0)). Among them, fliplr is a function to flip the matrix left and right along the vertical axis, and conj is a function to find the complex conjugate.

According to this, C ₁ =[-1 1], S ₁ =[-1 -1], and combine C ₀ C ₁ to generate a new complementary code as C ₀ '=[1 1-1 1], S ₀ ' =[1-1-1-1], at this time the length of each complementary code is expanded from 2 to 4.

Here, the length L _N of the complementary code can be designed (L _N is a power of 2), that is, the lengths of C _n and S _n are respectively L _N /2. Using the above method, the generated LAS code is iterated, and its length is extended to L _N , the number of iterations is log ₂ L _N -2, and the complementary codes generated finally are C _n and S _n .

Combine this pair of complementary codes and zero sequences to generate LAS codes, expressed in the form: Las=[C _n L ₀ S _n ], where L ₀ represents the number of 0s, that is, the shortest interval length between C _n and S _n , The length of the finally generated LAS code is expressed as L=L _N +L ₀ .

Fig. 8 shows the autocorrelation characteristic of the LAS code.

According to one aspect of the present invention, LAS codes are used to design training sequences.

For timing synchronization purposes, the training sequence includes at least one LAS code. Since the LAS short code still has a good synchronization effect in the case of a large frequency offset, preferably, the training sequence includes at least one LAS short code, represented by [X _las ] _SN , wherein the length of the LAS short code Denoted as SN, its complementary code length and zero sequence length are expressed as Lshort _-N and _Lshort-0 respectively, SN=Lshort _-N + _Lshort-0 .

In order to further optimize the autocorrelation characteristics of the LAS code, a zero sequence of the same length as the LAS short code may also be included before the LAS short code, represented by [0] _SN .

In a specific embodiment, the training sequence can include two identical LAS short codes, so that if one of the LAS short codes can be used for timing synchronization, it can also form a LAS short code pair with the other LAS short code for the carrier Synchronize.

For carrier synchronization purposes, the training sequence may include at least one pair of identical LAS codes. Since the LAS short codes still have a good synchronization effect in the case of a large frequency offset, preferably, the training sequence includes at least one pair of identical LAS short codes.

Preferably, the carrier synchronization can be divided into two stages, ie carrier coarse synchronization and carrier fine synchronization. Therefore, the training sequence may include at least two pairs of LAS codes. Preferably, one pair of LAS codes may be the same LAS short code for carrier coarse synchronization, and another pair of LAS codes may be the same LAS long code for carrier fine synchronization. The LAS long code can be represented by [X _las ] _LN , wherein the length of the LAS long code is denoted as LN, and its complementary code length and zero sequence length are expressed as L _length-N and L _length-0 respectively, LN=L _length-N + L _{length -0} .

In order to further optimize the cross-correlation characteristics of the LAS code, a zero sequence of the same length as the LAS short code may also be included before each LAS short code, represented by [0] _SN .

For the purpose of channel estimation, the training sequence may include at least one LAS code, such as a LAS long code, or may also include two LAS long codes, and perform channel estimation twice for these two long LAS codes, thereby improving the performance of channel estimation. Success rate.

As a specific example, it can be designed that Llong _-N =256, _Llong-0 =16; Lshort _-N =16, _Lshort-0 =8. Of course, the lengths of the LAS long code and the LAS short code here are only shown as examples, and other lengths can also be designed.

As a preferred embodiment, a LAS code training sequence that simultaneously satisfies timing synchronization, carrier synchronization and channel estimation can be designed as: [0] _SN , [X _las ] _SN , [0] _SN , [X _las ] _SN , [X _las ] _LN ,[X _las ] _LN . In this embodiment, the first LAS code is a short code, which can realize timing synchronization, and the LAS short code still has a good synchronization effect when the frequency deviation is large. The first and second LAS short codes can be used for carrier coarse synchronization. The advantage of short codes is that they can handle large frequency deviations. The last two LAS codes are long codes, which can be used for fine frequency offset correction and channel estimation.

Timing synchronization process

When the receiver receives the signal, it needs to be synchronized with the communication system first, including timing synchronization and carrier synchronization. The principle of timing synchronization is to directly calculate the autocorrelation operation between the received signal and the local LAS code through the matched filter method to obtain the autocorrelation peak value. Find the positions of the training symbols from the correlation peaks according to a certain method. Finding the position of the training symbol also determines the starting position of the current frame, that is, the time synchronization between the received signal and the system is completed, and the timing synchronization process ends.

As mentioned above, because the autocorrelation and cross-correlation characteristics of LAS codes are relatively good, LAS codes are used to design training symbols. Therefore, when calculating the correlation between the received signal and the LAS code, the peak size distribution is quite different. By setting the threshold reasonably, the starting position of the LAS code can be found very accurately, and the timing accuracy is high.

Specifically, when looking for the correlation peak value of the LAS code, according to the training symbol structure, the appropriate signal receiving length is adopted, and the sliding window autocorrelation operation method is used to correlate the received signal with the local LAS code to find the autocorrelation peak value to determine the LAS code. Location. For example, the received signal length here can be guaranteed to at least cover the LAS code, so as to ensure that the peak value can be detected.

The so-called sliding window autocorrelation operation uses the length of the LAS code as the window length to perform window processing on the received signal, and correlates the signal in the current window with the local LAS code to obtain an autocorrelation result. Then, the window is slid backwards, and then the received signal is windowed, and a correlation operation is performed between the signal in the current window and the local LAS code, so as to obtain another correlation result. In this way, the window is continuously slid until all received signals are correlated. From all the autocorrelation results calculated, the position of the LAS code is found by setting a threshold, that is, the autocorrelation results exceeding the threshold are taken as peak values.

In an example, only one LAS code is included in the training sequence, for example, one LAS short code, because the short code still has better synchronization effect under the condition of large frequency offset. In this case, the length of the LAS short code can be used as the window length to perform window processing on the received signal, and the signal in the current window is correlated with the local LAS short code to obtain an autocorrelation result. Then, the window is slid backwards, and then the received signal is windowed, and a correlation operation is performed between the signal in the current window and the local LAS code, so as to obtain another correlation result. In this way, the window is continuously slid until all received signals are correlated. From all the autocorrelation results calculated, the position of the LAS code is found by setting a threshold, that is, the autocorrelation results exceeding the threshold are taken as peak values.

In the case of a multi-path channel, it is possible that the amplitude of the next few paths is higher than that of the first path, and the first peak point exceeding the threshold should be selected, not necessarily the global maximum. Figure 9 shows a distribution plot of the autocorrelation results for timing synchronization. Suppose the threshold is 100, as shown in Figure 9, there are two autocorrelation results exceeding the threshold of 100, but the autocorrelation result at position 25 is selected as the peak value of this operation, so that the position at 25 is used as the found LAS code s position.

In the case of the previous preferred training symbol format [0] _SN ,[X _las ] _SN ,[0] _SN ,[X _las ] _SN ,[X _las ] _LN ,[X _las ] _LN , there exists in the training sequence Two LAS short codes. At this time, two peaks exceeding the threshold can be found through the above-mentioned sliding window autocorrelation calculation method. Fig. 9 shows a distribution diagram of autocorrelation results where there are two peaks. At this time, it is necessary to judge which one is the peak value of the previous short code and which one is the peak value of the rear short code.

Fig. 10 shows a schematic diagram of the training sequence in the case where two peaks are detected. In Fig. 10, two training sequences sent repeatedly in a loop are shown. The length of the received signal spans two training sequences, therefore, one of the two peaks found may be caused by the first LAS short code of the next training sequence. Therefore, it is necessary to determine which LAS short code each peak corresponds to.

Specifically, if the length of the interval between two peaks is 2*SN, then select the first peak that exceeds the threshold as the starting position of the first short LAS code, if the length of the interval between the two is greater than 2*SN, the second The first peak that exceeds the threshold is the starting position of the first short LAS code.

If there is a multi-path channel, then there will be two parts of concentrated distribution of correlation peaks after the sliding window, compare the correlation peaks of each part with the threshold value, select the first peak point that exceeds the threshold value, and after the comparison of the two parts, two parts will be obtained points exceeding the threshold, and then determine the position corresponding to the LAS code according to the above method.

In addition, if the transmitted signal passes through other band-limiting filters, the matched filtering results in smoother peaks rather than independent points, so the peak points need to be selected according to the actual band-limiting filter.

Fig. 11 shows a block diagram of a timing synchronization unit at a receiving end according to an aspect of the present invention. The timing synchronization unit may be part of the synchronization unit discussed above in connection with FIGS. 2 and 5 .

As shown in FIG. 11 , the timing synchronization unit 1100 may include an autocorrelation calculation unit 1110 for performing autocorrelation calculations. The autocorrelation calculation unit 1110 can take a window for the received signal, so as to use the local LAS code to perform autocorrelation calculation on the signal in the window, and slide the window to perform the next autocorrelation calculation until the signal receiving length is reached. The timing synchronization unit 1100 may further include a peak judging unit 1120 for judging the position of the peak according to the obtained correlation result set, so as to find the starting position of the LAS code. The peak determination unit 1120 may select an appropriate threshold, and use the autocorrelation result exceeding the threshold as the peak value.

Fig. 12 shows a flowchart of a timing synchronization method according to an aspect of the present invention. As shown, the method may include:

Step 1201: Take a window for the received signal, so as to use the local LAS code to perform autocorrelation calculation on the signal in the window, and slide the window to perform the next autocorrelation calculation until the signal reception length is reached; and

Step 1202: Determine the position of the peak according to the obtained correlation result set, so as to find the starting position of the LAS code.

As mentioned above, in the case of two LAS short codes, if the length of the interval between the two peaks is 2*SN, then select the first peak exceeding the threshold as the starting position of the first short LAS code, if both If the interval length is greater than 2*SN, then the second peak exceeding the threshold is the starting position of the first short LAS code.

Carrier Synchronization Process

After receiving the signal, it needs to be synchronized with the communication system first, including timing synchronization and carrier synchronization. The received signal and the system must first maintain time synchronization, obtain the starting position of the LAS code through timing synchronization, and then perform frequency synchronization.

For carrier synchronization, the training sequence information portion of the received signal includes at least one pair of identical LAS codes. The cross-correlation operation is performed on the repeated LAS codes to obtain the frequency deviation Δf.

Assuming that the carrier deviation between the receiver and the transmitter is Δf, and the AD sampling interval is T, then when the receiver ignores the influence of the noise signal, the received signal is expressed as:

y _n ＝x _n e ^j2πΔfnT

The correlation coefficient of the two LAS codes before and after is:

Where L represents the interval between LAS codes.

It can be seen from the above formula that the carrier frequency offset is:

Preferably, the training sequence information part may include two pairs of LAS codes, wherein a pair of identical LAS codes are LAS short codes, so that carrier coarse synchronization can be performed first; in addition, a pair of identical LAS long codes is included, thereby Carrier fine synchronization is possible.

Since the timing synchronization has been completed, the corresponding two parts of the short LAS code can be extracted according to the training symbol index returned by the timing synchronization, and the carrier of the short LAS code is roughly synchronized. The short code can handle a large frequency offset. Calculated according to the above formula The estimated frequency offset value is Δf ₁ . Then extract two parts of the long LAS code, correct the fine frequency offset of the carrier for the long LAS code, and obtain the estimated frequency offset value Δf ₂ , refer to the frequency offset of the coarse synchronization, the final output frequency offset is Δf=Δf ₁ + Δf ₂ .

Take the previous better training symbol format [0] _SN , [X _las ] _SN , [0] _SN , [X _las ] _SN , [X _las ] _LN , [X _las ] _LN as an example. Let LN=272, SN=24, and the total length of training symbols is 640. The two short LAS codes are located at (25:48) and (73:96) respectively, and the long LAS codes are located at (97:368) and (369:640) respectively.

Ideally, the start position of the LAS code obtained through timing synchronization calculation is the start position of the first short LAS code, which is 25. According to the index and the code lengths LN and SN of the long and short codes, corresponding codes are correspondingly extracted from the received signal.

Carrier Coarse Synchronization

Two parts of short LAS codes are extracted from the received signal, according to the formula Multiply its conjugates to get the correlation coefficient R. Then according to the formula Calculate the corresponding coarse frequency offset Δf ₁ , where L represents the interval between two short LAS codes, and it can be seen from the training symbol structure that L=2*SN=48.

According to the calculated coarse frequency offset by the formula Frequency offset correction is performed on the received signal to obtain a signal after the first frequency offset correction.

Carrier Fine Frequency Offset Correction

In coarse carrier synchronization, a coarse frequency offset correction is performed on the received signal to obtain the received signal y _n '. The process of fine frequency offset is to extract two parts of long LAS code from y _n ', according to the formula Multiply its conjugates to get the correlation coefficient R. Then according to the formula Calculate the corresponding fine frequency offset Δf ₂ , L represents the interval between two long LAS codes, it can be seen from the training symbol structure, L=LN=272.

Referring to the frequency offset of the coarse synchronization, the final output frequency offset is Δf=Δf ₁ +Δf ₂ . And according to the formula y _n ”=y _n 'e ^j2π(-Δf)nT , the signal after correcting the fine frequency offset of the received signal is obtained.

The signal y _n ” after twice frequency offset correction is used as an input signal to the channel estimation process, and the carrier synchronization process ends.

FIG. 13 shows a block diagram of a carrier synchronization unit 1300 . The carrier synchronization unit 1300 may be part of the synchronization unit discussed above in connection with FIGS. 2 and 5 .

As shown in the figure, the carrier synchronization unit 1300 may include a cross-correlation calculation unit 1310 and a frequency correction unit 1320 . The cross-correlation calculation unit 1310 can perform cross-correlation calculation on a pair of LAS codes to obtain the frequency offset of the carrier between the receiving end and the transmitting end. The frequency correction unit 1320 may perform frequency offset correction on the received signal according to the frequency offset of the carrier.

In an embodiment, the cross-correlation calculation unit 1310 may first perform a cross-correlation calculation on a pair of LAS short codes to obtain a coarse frequency offset of the carrier between the receiving end and the transmitting end. The frequency correction unit 1320 may first perform a primary frequency offset correction on the received signal according to the rough frequency offset. The cross-correlation calculation unit 1310 performs cross-correlation calculation on a pair of LAS long codes extracted from the received signal after primary frequency offset correction to obtain a fine frequency offset of the carrier between the receiving end and the transmitting end. The frequency correction unit 1320 may then perform a second frequency offset correction on the received signal after primary frequency offset correction according to the fine frequency offset and the coarse frequency offset, so as to obtain a final frequency offset corrected signal.

Fig. 14 shows a flowchart of a carrier synchronization method according to an embodiment. As shown in the figure, the carrier synchronization method may include the following steps:

Step 1401: Perform cross-correlation on the two LAS codes extracted from the received signal to obtain the frequency offset of the carrier between the receiving end and the transmitting end; and

Step 1402: Perform frequency offset correction on the received signal based on the frequency offset.

Fig. 15 shows a flowchart of a carrier synchronization method according to another embodiment. As shown in the figure, the carrier synchronization method may include the following steps:

Step 1501: Perform cross-correlation on the two LAS short codes extracted from the received signal to obtain a coarse frequency offset of the carrier between the receiving end and the transmitting end;

Step 1502: Perform primary frequency offset correction on the received signal according to the coarse frequency offset;

Step 1503: Perform cross-correlation calculation on a pair of LAS long codes extracted from the received signal after primary frequency offset correction to obtain a fine frequency offset of the carrier between the receiving end and the transmitting end; and

Step 1504: According to the fine frequency offset and the coarse frequency offset, perform a second frequency offset correction on the received signal after the primary frequency offset correction.

Although the methods described above are illustrated and described as a series of acts for simplicity of explanation, it is to be understood and appreciated that the methodologies are not limited by the order of the acts, as some acts may occur in a different order according to one or more embodiments And/or concurrently with other actions from those illustrated and described herein or not illustrated and described herein but can be understood by those skilled in the art.

channel estimation process

Channel estimation is used to estimate the transmission characteristics of the channel, ie the influence of the channel on the transmitted signal. By utilizing the training symbols known by both the transmitting end and the receiving end, the receiving end can perform channel estimation according to the known training symbols and the received training symbols. For example, the receiving end can perform correlation on known training symbols and received training symbols, so as to determine the transmission characteristic of the channel. After performing channel estimation, the receiving end can use the determined channel estimation to demodulate the received unknown data signal, so as to determine the actual data signal sent by the transmitting end.

The received signal is timed and synchronized with the system to maintain time synchronization. Then perform carrier synchronization with the received signal. Carrier synchronization includes coarse synchronization and fine synchronization. The carrier frequency offset Δf of the receiver and the transmitter is obtained through synchronization, and the received signal is corrected through the carrier frequency offset to obtain the corrected received signal. y _fix , do channel estimation on y _fix .

The present invention uses the LAS code as the training sequence, for example, the long LAS code L-LAS in the training symbol format can be used for channel estimation.

Channel estimation can be expressed as:

Among them, y _n represents the received signal corrected by carrier synchronization, that is, y _fix . N represents the length of the LAS code. x _n represents a local LAS code, that is, x _n represents one of the last two long LAS codes in the training symbols. R ₀ represents the sum of squares of the LAS code, and P represents the number of multipath channels.

The channel estimator estimates the impulse response h(t) of the channel from the received signal _yfix of the training symbols, and then constructs an inverse channel system according to the estimated h(t), and the received data signal is passed through the inverse channel system by Reverts to an estimate of the signal fed to the channel by the sender.

The general received signal y _n can be expressed as _en means noise. After substituting it into the above formula, the following formula is obtained:

Represents the autocorrelation of the training sequence. By rationally designing the autocorrelation coefficient to be zero, the estimated channel is highly close to the real channel, thereby greatly improving the accuracy of channel estimation. According to the present invention, since the autocorrelation of the LAS code has a very high probability of 0, the success rate of channel estimation is greatly improved during channel estimation.

In the field, M sequences are generally used for channel estimation. The autocorrelation characteristics of the M sequence are shown in Figure 7. From the figure, it can be seen that the autocorrelation characteristics will appear pulses at a certain time interval, and its autocorrelation characteristics are not very good. The corresponding channel estimation formula

middle The probability that the value is not 0 is very high, so the estimated channel model deviates greatly from the ideal channel model, which has a great impact on subsequent decoding processing and increases the bit error rate of the system.

Compared with the LAS code sequence, it has the characteristics that the autocorrelation function is an ideal impulse function at the origin, zero everywhere outside the origin, and the cross-correlation function is zero everywhere. Therefore, when doing channel estimation, the actual estimated channel model and the ideal model The deviation is very small, which reduces the bit error rate of the system and improves the system performance very well.

According to the present invention, since there are two long LAS codes in the training symbols, the channel estimation process can be realized by using any one of the long LAS codes, or can also perform channel estimation twice for the two long LAS codes, thereby improving the channel estimation. success rate.

There may be one channel or multi-path channels in the communication environment, and the receiver can determine whether there are multi-path channels according to the environment. In the case of no multipath channel, that is, p=0, the channel estimate h can be directly calculated according to the above formula. In the case of multi-path channels, the channel estimation value h _p of each multi-path path can be calculated separately according to the above formula, where the local LAS code x _n is offset for each multi-path path, and the deviation of each path Can be 1.

For example, there may be, for example, 6 actual multipath channels. Firstly, the local LAS codes are arranged into 6 columns according to the number of multipaths, and the path deviation of each column is 1. The arrangement is shown in Figure 16.

According to the training symbol format [0] _SN ,[X _las ] _SN ,[0] _SN ,[X _las ] _SN ,[X _las ] _LN ,[X _las ] _LN , find the corresponding LAS code position from the correction signal y _fix , and extracted as y _fix-las , a total of two parts.

Combine the extracted y _fix-las with the local LAS codes of the rearranged 6 multipath channels through the formula

After processing, the channel estimation value h _p of each multipath path is obtained. Since there are two parts of LAS codes that can perform channel estimation, each part will obtain channel estimation value h _p after processing, and the final channel estimation value h _p of each multipath path can be obtained by averaging the two parts.

Then, the received data signal can be demodulated based on the channel estimation value _hp of each multipath path, so as to recover the signal of the transmitting end of each multipath path.

Design training sequence bandwidth

The design symbol structure in this system includes training sequence TSC (traning sequence code) and data (data). The design of training symbols is very important, which affects the three most important links of timing, synchronization, and channel estimation of the entire system. If any of these three steps has a large error, it will have a great impact on the entire system. Follow-up The decoding process is meaningless.

The design process of the bandwidth of the training sequence is relatively complicated. When the bandwidth is short, the corresponding power spectral density is relatively large. When there are multiple carriers in the system, it will affect the reception and transmission of data. When the bandwidth is too large, the corresponding power spectral density If it is too small, the sensitivity of the system's transmitter and receiver is extremely high.

In the existing communication system, the method of using the same bandwidth of the training sequence and data is generally used, and the corresponding power spectral density is the same, and because the bandwidth of the general system is relatively short, it corresponds to a long time domain transmission time, which affects the signal The processing time of synchronization and channel estimation, and the waiting time of the subsequent decoding process also become longer, which reduces the transmission rate of the system. In addition, since the training sequence takes a long time to send, when the signal is sampled, the sampling rate is low, and the time resolution is not fine enough, which affects the deviation of the channel estimation.

The present invention makes the training sequence bandwidth far greater than the data bandwidth (for example, 5 times, 10 times, 15 times or more), thereby the power spectral density of the training sequence is lower than the power spectral density of the data, and the bandwidth of its training sequence and data and power spectral density relationship diagram as shown in Figure 17. Since the transmission power of the training sequence and data needs to be consistent, it can be seen from the figure that when the bandwidth of the training sequence is widened, the corresponding power spectral density will also be greatly reduced. Compared with the power spectral density of the data Words are very low.

This system can use all available spreading codes, including m-sequence, Golomb code, CAN (Cyclic Algorithm New), and LAS code. In this system, we take the LAS code with complete complementary orthogonal characteristics as an example to introduce the processing process of timing synchronization, carrier synchronization and channel estimation. Therefore, all methods and devices described above using LAS codes as training codes for timing synchronization, carrier synchronization, and channel estimation are also applicable to all suitable spreading codes as training codes for timing synchronization, carrier synchronization, and training estimation. Therefore, the timing synchronization, carrier synchronization and channel estimation algorithms shown above with the LAS code as an example are only shown as examples, and the above content of the present invention is applicable to all suitable training codes.

The characteristic of the LAS code is that the autocorrelation function is an ideal impact function at the origin, and it is zero everywhere outside the origin, while the cross-correlation function is zero everywhere. The autocorrelation characteristics of the LAS code are shown in Figure 8. Therefore, when the training sequences overlap, they will not interfere with each other. This design can improve the spectrum utilization and transmission rate of the system.

by the formula It can be seen that when the bandwidth in the frequency domain is larger, the corresponding time in the time domain is smaller, that is, the process of sending and receiving the training sequence can be completed in a shorter time. In the signal receiving process, for the same length of data, when the receiving time becomes shorter, the sampling rate of the signal can be increased to make the time resolution finer. In the channel estimation process, the accuracy of the time resolution is improved, so that the channel estimation result is more accurate.

On the one hand, since the power spectral density of the training sequence is extremely low, it hardly affects the data signal, so the training sequence and data can be superimposed and sent at the same time. In other words, the training sequence and data are transmitted at least partially overlapping in frequency and/or time. When there are two carrier signals sending data at the same time, its structure diagram is shown in Figure 18. It can be seen from the figure that there is a guard band between the actual data carried by the two carrier signals, which will not overlap or cause mutual interference ; and the frequency width of the training sequence overlaps with the actual data, because the power spectral density of the training sequence is very low, so it will not cause interference to the actual data; moreover, different training sequences can be distinguished by different spreading codes, which will not cause confused. The training sequence does not occupy specific frequency and time resources, which improves the spectrum utilization and transmission rate of the system.

In one embodiment, the LAS code with complete complementary orthogonal characteristics can be used as the training sequence in this system. It is characterized in that the autocorrelation function is an ideal impact function at the origin, and it is zero everywhere outside the origin, and the cross-correlation function is everywhere. Zero, the autocorrelation and cross-correlation characteristics of the LAS code are shown in Figure 5. Therefore, when the training sequences overlap, they will not interfere with each other. This design can improve the spectrum utilization and transmission rate of the system.

In this case, we design the format of the training sequence as: [0] _SN ,[X _las ] _SN ,[0] _SN ,[X _las ] _SN ,[X _las ] _LN ,[X _las ] _LN .

Those of skill in the art would understand that information, signals and data may be represented using any of a variety of different technologies and techniques. For example, data, instructions, commands, information, signals, bits, symbols, and chips referenced throughout the above description may be composed of voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields or particles, or any combination thereof. To represent.

Those of skill in the art would further appreciate that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

The various illustrative logic modules, and circuits described in connection with the embodiments disclosed herein may be implemented using a general-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable Logic devices, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein are implemented or performed. A general-purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in cooperation with a DSP core, or any other such configuration.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of both. A software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor such that the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integrated into the processor. The processor and storage medium can reside in an ASIC. The ASIC may reside in a user terminal. In the alternative, the processor and storage medium may reside as discrete components in the user terminal.

In one or more exemplary embodiments, the functions described may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, as a computer program product, the functions may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a computer. By way of example and not limitation, such computer-readable media may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or other Any other medium that is suitable for program code and can be accessed by a computer. Any connection is also properly termed a computer-readable medium. For example, if the software is transmitted from a web site, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave , then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of media. Disk and disc, as used herein, include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk, and Blu-ray disc, where disks are often reproduced magnetically. data, while a disc (disc) uses laser light to reproduce data optically. Combinations of the above should also be included within the scope of computer-readable media.

The previous description of the present disclosure is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to the present disclosure will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other variations without departing from the spirit or scope of the present disclosure. Thus, the disclosure is not intended to be limited to the examples and designs described herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (18)

1. A carrier synchronization method, comprising: Perform a cross-correlation operation on two training codes extracted from the received signal to obtain the frequency offset of the carrier between the receiving end and the transmitting end, wherein the received signal is generated from the initial frequency domain data signal after Fourier inverse transform The data signal and the training sequence signal in the time domain are composed of a training sequence signal in the time domain, and the training sequence signal in the time domain includes a training sequence based on a training code, wherein the training sequence bandwidth is greater than the data bandwidth and the power spectral density of the training sequence is lower than the power spectral density of the data ;as well as Frequency offset correction is performed on the received signal based on the frequency offset. 2. The carrier synchronization method according to claim 1, wherein the training code comprises m-sequence, Golomb code, CAN code, or LAS code. 3. The carrier synchronization method according to claim 1, wherein the training sequence bandwidth is 5 times, 10 times, 15 times or more than the data bandwidth. 4. carrier synchronization method as claimed in claim 1, is characterized in that, described training sequence comprises two LAS short codes [X _las ] _SN , SN is the length of described LAS short code, and wherein, described pair is extracted from The two training codes of the received signal perform a cross-correlation operation to obtain the frequency offset of the carrier between the receiving end and the transmitting end including: performing a cross-correlation operation on the two LAS short codes extracted from the received signal to obtain a coarse frequency offset of the carrier between the receiving end and the transmitting end, and The performing frequency offset correction on the received signal based on the frequency offset includes: performing a primary frequency offset correction on the received signal based on the coarse frequency offset. 5. The carrier synchronization method according to claim 4, wherein the training sequence comprises [0] _SN , [X _las ] _SN , [0] _SN , [X _las ] _SN , wherein [0] _SN is A sequence of 0s of length SN. 6. carrier synchronization method as claimed in claim 4, is characterized in that, described training sequence also comprises two LAS long codes [X _las ] _LN , LN is the length of described LAS long code, wherein, described to extract Performing a cross-correlation operation on two training codes from the received signal to obtain the frequency offset of the carrier between the receiver and the transmitter also includes: performing a cross-correlation operation on two LAS long codes extracted from the received signal subjected to the primary frequency offset correction to obtain a fine frequency offset of the carrier between the receiving end and the transmitting end, and The performing frequency offset correction on the received signal based on the frequency offset further includes: Performing secondary frequency offset correction on the received signal after the primary frequency offset correction based on the coarse frequency offset and the fine frequency offset. 7. The carrier synchronization method according to claim 6, wherein said performing secondary frequency offset correction on the received signal after said primary frequency offset correction based on said coarse frequency offset and said fine frequency offset comprises: The secondary frequency offset correction is performed on the received signal after the primary frequency offset correction based on the sum of the coarse frequency offset and the fine frequency offset. 8. The carrier synchronization method according to claim 6, wherein the training sequence comprises [0] _SN , [X _las ] _SN , [0] _SN , [X _las ] _SN , [X _las ] _LN , [X _las ] _LN , where [0] _SN is a 0-sequence whose length is SN. 9. The carrier synchronization method according to claim 6, wherein the two LAS short codes are also used for timing synchronization, and the extraction positions of the two LAS short codes and the two LAS long codes are Determined based on timing synchronization results. 10. A carrier synchronization device, comprising: A cross-correlation calculation unit, configured to perform a cross-correlation operation on two training codes extracted from the received signal, wherein the received signal is obtained by Fourier transforming the initial frequency domain data signal to obtain the frequency offset of the carrier between the receiving end and the transmitting end. The time-domain data signal generated after the inverse transformation is composed of a time-domain training sequence signal and the time-domain training sequence signal includes a training sequence based on a training code, wherein the training sequence bandwidth is greater than the data bandwidth and the power spectral density of the training sequence is below the power spectral density of the data; and A frequency correction unit, configured to perform frequency offset correction on the received signal based on the frequency offset. 11. The carrier synchronization device according to claim 10, wherein the training code comprises m-sequence, Golomb code, CAN code, or LAS code. 12. The carrier synchronization device according to claim 10, wherein the training sequence bandwidth is 5 times, 10 times, 15 times or more than the data bandwidth. 13. The carrier synchronization device according to claim 10, wherein the training sequence comprises two LAS short codes [X _las ] _SN , and SN is the length of the LAS short code, wherein the correlation calculation unit It is further used to perform a cross-correlation operation on the two LAS short codes extracted from the received signal, so as to obtain a coarse frequency offset of the carrier between the receiving end and the transmitting end, and The frequency correction unit is further configured to perform primary frequency offset correction on the received signal based on the coarse frequency offset. 14. The carrier synchronization device according to claim 13, wherein the training sequence comprises [0] _SN , [X _las ] _SN , [0] _SN , [X _las ] _SN , wherein [0] _SN is A sequence of 0s of length SN. 15. The carrier synchronization device according to claim 13, wherein the training sequence further comprises two LAS long codes [X _las ] _LN , and LN is the length of the LAS long code, wherein the correlation calculation The unit is further configured to perform a cross-correlation operation on the two LAS long codes extracted from the received signal after the initial frequency offset correction, so as to obtain a fine frequency offset of the carrier between the receiving end and the transmitting end, and The frequency correction unit is further configured to perform a secondary frequency offset correction on the received signal after the primary frequency offset correction based on the coarse frequency offset and the fine frequency offset. 16. The carrier synchronization device according to claim 15, wherein the frequency correction unit is further configured to correct the received frequency offset after the initial frequency offset correction based on the sum of the coarse frequency offset and the fine frequency offset The signal performs the secondary frequency offset correction. 17. The carrier synchronization device according to claim 15, wherein the training sequence comprises [0] _SN , [X _las ] _SN , [0] _SN , [X _las ] _SN , [X _las ] _LN , [X _las ] _LN , where [0] _SN is a 0-sequence whose length is SN. 18. The carrier synchronization device according to claim 15, wherein the two LAS short codes are also used for timing synchronization, and the extraction positions of the two LAS short codes and the two LAS long codes are Determined based on timing synchronization results.