Symbol synchronization and frequency offset estimation method and system in OFDM system based on CHU sequence

Abstract

The invention discloses a symbol synchronization and frequency offset estimation method and system in an OFDM system based on a CHU sequence, and the method comprises the steps: a transmitting end obtains a conjugate symmetric original sequence based on the splicing of the CHU sequence, and generates a conjugate antisymmetric training sequence based on the conjugate symmetric original sequence; adding the training sequence into a data frame for transmitting a data signal, and transmitting the training sequence in a data transmission channel; a receiving end receives a data signal comprising a training sequence, calculates correlation values of different data positions of the signal by respectively adopting a Schmidl strategy, a Minn strategy and a symmetric structure training sequence synchronization strategy, calculates corresponding timing metric values based on the correlation values, and calculates a comprehensive timing metric value based on the timing metric values calculated by the strategies; and a receiving end performs symbol synchronization in the OFDM system based on the comprehensive timing metric value, and performs frequency offset estimation based on a correlation value of a symbol synchronization position. According to the invention, symbol and frequency joint synchronization with a high main peak value, a low auxiliary peak value and no peak platform problem can be realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of OFDM system synchronization and frequency offset estimation, and relates to a method and system for symbol synchronization and frequency offset estimation in an OFDM system based on a CHU sequence. Background Art

[0002] The OFDM (Orthogonal Frequency Division Multiplexing) system has attracted widespread attention due to its ability to effectively combat frequency selective fading and narrowband interference and its high spectrum resource utilization rate. It is the core technology of the communication system. OFDM technology is a type of multi-carrier modulation technology that converts high-speed data signals into low-speed sub-data streams and modulates them to each sub-channel for transmission. The orthogonality between the sub-carriers in the system allows the sub-channel spectra to overlap with each other, thereby improving the spectrum resource utilization rate.

[0003] However, OFDM systems are very sensitive to symbol timing errors and frequency offsets. OFDM symbols composed of multiple subcarriers are easily affected by synchronization errors. Symbol timing synchronization errors cause inter-symbol interference. Carrier frequency offset destroys the orthogonality between subcarriers, causing inter-carrier interference and reducing system performance. Therefore, the receiver must achieve high-precision symbol and frequency synchronization. Symbol timing and carrier frequency synchronization have become important technologies in OFDM systems.

[0004] Synchronization technology is a problem that needs to be solved in any communication system. The performance of the synchronization algorithm directly affects the performance of the entire communication system. It can be said that without an accurate synchronization algorithm, reliable data transmission cannot be achieved. The performance of the synchronization algorithm directly affects whether the receiving end can accurately demodulate and restore the original data. The role of symbol timing synchronization includes the receiving end estimating the position of the OFDM data frame through the synchronization algorithm and finding the precise position of the FFT window. The existing symbol timing synchronization algorithms are divided into autocorrelation and cross-correlation synchronization algorithms. The cross-correlation timing synchronization method is implemented by pre-sharing local sequences between the transmitter and the receiver. It has good anti-noise performance and can obtain good timing synchronization, but it is very sensitive to carrier frequency deviation. The autocorrelation timing synchronization method is not affected by the frequency deviation because the data sequences with repeated or conjugate structures participating in the correlation operation experience the same frequency deviation, but the synchronization performance is greatly affected by noise. Due to the complex hardware implementation of the iterative operation of the cross-correlation timing synchronization method, the existing communication system mostly uses the autocorrelation symbol frequency synchronization algorithm based on the training sequence.

[0005] The carrier frequency offset originates from the inconsistency between the signal laser at the transmitting end and the local oscillator frequency at the receiving end. The deviation of the carrier frequency between the transmitter and the receiver is a key factor affecting the performance of the OFDM system. In a single-carrier communication system, the symbol synchronization performance affects the system transmission performance. The carrier frequency offset will only cause a certain attenuation and phase rotation to the received signal, which can be solved by equalization and other methods. For OFDM multi-carrier systems that require strict orthogonality between subcarriers, carrier offset will cause inter-carrier interference (ICI), which will have a serious impact on system performance. Due to the loss of orthogonality between subcarriers, increasing the signal transmission power cannot significantly improve the performance of the system. Therefore, both single-carrier communication systems and OFDM communication systems require accurate synchronization. Classic symbol and frequency joint synchronization algorithms include the Schmidl algorithm, the Minn algorithm, the Park algorithm, etc. In order to achieve symbol and frequency synchronization of the OFDM system, the OFDM frame structure based on the autocorrelation symbol and frequency synchronization algorithm of the training sequence is as follows Figure 1 The Schmidl algorithm implements symbol frequency synchronization based on the training sequence. The training sequence structure is as follows: Figure 2 As shown. This training sequence is composed of two identical sequences of length N / 2 in the time domain. The receiving end of the Schmidl algorithm calculates the correlation value of the OFDM signal by using a sliding window of length L. This window can slide over time to calculate the autocorrelation values ​​and timing metrics of data signals at different positions. The symbol timing metric method proposed by Schmidl is obtained by calculating the correlation values ​​of the left N / 2 sequence and the right N / 2 sequence of the training sequence. When the position of the sliding window is within the cyclic prefix, the timing metric obtains the maximum value, so a "platform" will appear, which will lead to inaccurate judgment of the timing synchronization position. Based on the training sequence with a repetitive structure, the Schmidll algorithm can use the autocorrelation value of the symbol synchronization position to estimate the fractional part of the frequency offset (FO), but the use of the Schmidll algorithm may result in a large error in the frequency offset estimation. The method for estimating the frequency offset based on the autocorrelation value of the timing synchronization position of the Schmidl algorithm is as shown in formula (1):

[0006]

[0007] is the frequency deviation value, angle(P Sch (n)) is the autocorrelation value P of the synchronization position n Sch (n) angle.

[0008] Schmidl's symbol timing synchronization strategy has good reliability under low signal-to-noise ratio (SNR), but due to the existence of the peak platform of the timing metric, it may lead to a decrease in the accuracy of symbol frequency synchronization. Minn solves the peak platform problem by modifying the structure of the training sequence.

[0009] The training sequence structure used by the Minn algorithm is [A AAA], such as Figure 3 As shown, A represents a sequence with a length of L=N / 4, which is generated by IFFT modulation of an N / 4-point PN sequence. The method proposed by Minn also obtains timing measurement by calculating the autocorrelation value of the training sequence. The Minn algorithm uses pseudo-random sequences with opposite relationships, which can make the result of the timing measurement form a maximum value, solving the peak platform problem of the Schmidl algorithm. However, due to the particularity of the training sequence structure, secondary peak interference is caused, which may cause misjudgment of the timing position. In addition, when the signal-to-noise ratio decreases and the training sequence is interfered, the secondary peak peak value in the timing measurement will be close to the main peak peak value. The Minn algorithm solves the peak platform problem, but the secondary peak interferes with the main peak, resulting in errors in symbol synchronization, and the frequency offset estimation based on the autocorrelation value of the symbol synchronization position also has deviations. The method for estimating the frequency offset based on the autocorrelation value of the timing synchronization position based on the Minn algorithm is as shown in formula (2),

[0010]

[0011] is the frequency deviation value, angle(P Minn (n)) is the autocorrelation value P of the synchronization position n Minn (n) angle.

[0012] In order to solve the above problems, Park et al. proposed a new symbol timing synchronization algorithm and designed a new training sequence structure, such as Figure 4 As shown. The new training sequence structure consists of 4 parts. C is the complex PN sequence obtained by IFFT modulation, and its length is N / 4; C* is the conjugate sequence of C, and its length is N / 4; D is the symmetric sequence of C, and its length is also N / 4. The left and right halves of the Park algorithm training sequence structure are symmetric sequences at their respective midpoints. Due to the structure of the training sequence, the correlation value drops rapidly when the sliding window deviates from the correct position. The Park algorithm solves the problem of the Minn algorithm's secondary peak. When the signal-to-noise ratio decreases, the synchronization accuracy decreases, resulting in errors in symbol timing and corresponding frequency offset estimation.

[0013] In summary, among the synchronization algorithms of existing OFDM systems, the Schmidl algorithm is not accurate enough in judging the synchronization position due to the peak platform problem, which may affect the frequency offset estimation using the autocorrelation value of the timing measurement peak position, resulting in deviations in the frequency offset estimation. The Minn algorithm solves the peak platform problem, but there is a problem of high secondary peak value. Due to the interference of the secondary peak, the synchronization position may be misjudged, resulting in errors in the subsequent frequency offset estimation, affecting the accuracy of the frequency offset estimation. The main peak of the Park algorithm is lower, there is no peak platform problem, and the problem of high secondary peak value of the Minn algorithm is solved. However, when the OSNR (optical signal-to-noise ratio) decreases, the peak value of the main peak of the Park algorithm's timing measurement decreases rapidly, resulting in a decrease in judgment accuracy and errors in symbol frequency synchronization.

[0014] Prior art document 1 (CN113162882A) provides an autocorrelation OFDM symbol synchronization method based on a conjugate antisymmetric training sequence. By using the mirror sequence, conjugate sequence and opposite sequence of the complex PN sequence to set the training sequence structure, the symbol timing synchronization accuracy in the OFDM system under high OSNR scenarios is significantly improved. However, the synchronization accuracy is low under low OSNR scenarios. Summary of the invention

[0015] In order to solve the deficiencies in the prior art, the present invention provides a method and system for symbol synchronization and frequency offset estimation in an OFDM system based on a CHU sequence. Constant amplitude zero autocorrelation sequence, especially the CHU sequence, has obvious constant amplitude, ideal autocorrelation and good cross-correlation characteristics. The present invention is upgraded from the aspects of training sequence structure, symbol synchronization timing measurement and frequency offset estimation, fully extracts the characteristics of the training sequence, improves the accuracy of symbol synchronization and frequency offset estimation in the OFDM system, realizes the joint synchronization of symbols and frequencies with high main peak peak value, low secondary peak peak value and no peak platform problem, and can accurately perform symbol synchronization and frequency offset estimation in wireless communication systems and optical communication systems.

[0016] The present invention adopts the following technical solution.

[0017] The first aspect of the present invention proposes a method for symbol synchronization and frequency offset estimation in an OFDM system based on a CHU sequence, comprising:

[0018] The transmitter obtains a conjugate symmetric original sequence U based on the CHU sequence concatenation, and generates a conjugate antisymmetric training sequence based on the conjugate symmetric original sequence U;

[0019] Adding the training sequence to a data frame of a signal in an OFDM system, wherein the data signal including the training sequence is transmitted in a data transmission channel;

[0020] The receiving end receives a data signal including a training sequence, and respectively uses a Schmidl strategy, a Minn strategy, and a symmetric structure training sequence synchronization strategy to calculate correlation values ​​of different data positions of the data signal based on the training sequence structure, calculates corresponding timing metric values ​​based on the correlation values, and calculates a comprehensive timing metric value based on the timing metric values ​​calculated by each strategy;

[0021] The receiving end performs symbol synchronization of the OFDM system based on the comprehensive timing measurement value, and performs frequency offset estimation based on the correlation value of the symbol synchronization position.

[0022] Preferably, the expression for obtaining the conjugated symmetric original sequence U based on CHU sequence splicing is:

[0023] U=[CHU conj(CHU')] (3)

[0024] Wherein, U represents the conjugate symmetric original sequence of length N / 8, CHU represents the CHU sequence of N / 16 points, N represents the training sequence length, CHU' represents the symmetric sequence of the CHU sequence, and conj represents the conjugate of the sequence;

[0025] The lth variable in the CHU sequence is:

[0026]

[0027] Wherein, L is the length of the CHU sequence; r is an integer coprime with L; l = 0, 1…, L-1.

[0028] Preferably, a conjugate antisymmetric training sequence is generated based on the conjugate symmetric original sequence U:

[0029] [-U,-U,-U*,U*,-U*,U*,U,U];

[0030] Among them, -U is the opposite number sequence of U, U* is the conjugate sequence of U, and -U* is the conjugate opposite number sequence of U.

[0031] Preferably, the Schmidt strategy uses formula (8) to add a negative number and a conjugate operator to the training sequence position corresponding to the nth data position of the received data signal, and calculates the correlation value P of the sequence obtained by adding the negative number and the conjugate operator based on the repetitive structure of [U,U,U*,U*,U,U,U*,U*] Sch (n), and calculate the timing metric value M by combining formulas (9) and (10) Sch (n):

[0032]

[0033]

[0034]

[0035] in,

[0036] P Sch (n) represents the correlation value of the nth data position in the data signal obtained using the Schmidl strategy;

[0037] Uc(m) and Uc(m+N / 2) are the m-th and m+N / 2-th data of the pre-shared local sequence [U,U,U*,U*,U*,U*,U,U];

[0038] r(n+m) and r(n+m+N / 2) respectively represent the n+m and n+m+N / 2th data in the training sequence position corresponding to the nth data position in the data signal received by the receiving end;

[0039] R Sch (n) is used for P Sch (n) coefficient for normalization;

[0040] M Sch (n) represents the timing metric value of the nth data position in the data signal obtained using the Schmidl strategy;

[0041] m is the relative position variable relative to n.

[0042] Preferably, the Minn strategy uses formula (11) to add a negative number and a conjugate operator to the training sequence position corresponding to the nth data position of the received data signal, and calculates the correlation value P of the sequence obtained by adding the negative number and the conjugate operator based on the cyclic repetition structure of [U,U,-U*,-U*,U,U,-U*,-U*] Minn (n), and calculate the timing metric value M by combining formulas (12) and (13) Minn (n):

[0043]

[0044]

[0045]

[0046] Among them, P Minn (n) represents the correlation value of the nth data position in the data signal obtained by using the Minn strategy;

[0047] Uc(m), Uc(m+(k-1)N / 8), Uc(m+N / 4+(k-1)N / 8), Uc(m+N / 2+(k-1)N / 8), and Uc(m+3N / 4+(k-1)N / 8) are the m, m+(k-1)N / 8, m+N / 4+(k-1)N / 8, m+N / 2+(k-1)N / 8, and m+3N / 4+(k-1)N / 8 data of the pre-shared local sequence [U,U,U*,U*,U*,U*,U,U];

[0048] r(n+m), r(n+m+(k-1)N / 8), r(n+m+N / 4+(k-1)N / 8), r(n+m+N / 2+(k-1)N / 8), and r(n+m+3N / 4+(k-1)N / 8) respectively represent the n+m, n+m+(k-1)N / 8, n+m+N / 4+(k-1)N / 8, and n+m+3N / 4+(k-1)N / 8 data in the training sequence position corresponding to the nth data position in the data signal received by the receiving end;

[0049] R Minn (n) is used for P Minn (n) coefficient for normalization;

[0050] M Minn (n) represents the timing metric value of the nth data position in the data signal obtained using the Minn strategy;

[0051] k is the control variable and m is the relative position variable relative to n.

[0052] Preferably, the symmetric structure training sequence synchronization strategy uses formula (5) to calculate the correlation value P of the sequence in the training sequence position corresponding to the nth data position of the received data signal based on the antisymmetric structure of [-U,-U,-U*,U*,-U*,U*,U,U] Sym (n), and calculate the timing metric value M by combining formulas (6) and (7) Sym (n):

[0053]

[0054]

[0055]

[0056] Among them, P Sym (n) represents the correlation value of the nth data position in the data signal obtained by using the symmetric structure training sequence synchronization strategy;

[0057] Uc(m), Uc(N / 2+1-m), and Uc(m+N / 2) are the mth, N / 2+1-mth, and m+N / 2th data of the local sequence [U,U,U*,U*,U*,U*,U] pre-shared between the transmitter and the receiver;

[0058] r(n+1-m) and r(n+m) respectively represent the n+1-mth and n+mth data in the training sequence position corresponding to the nth data position in the data signal received by the receiving end;

[0059] R Sym (n) is used for P Sym (n) coefficient for normalization;

[0060] M Sym (n) represents the timing metric value of the nth data position in the data signal obtained by using the symmetric structure training sequence synchronization strategy;

[0061] k is the control variable and m is the relative position variable relative to n.

[0062] Preferably, the formula for calculating the comprehensive timing metric value based on the timing metric values ​​calculated by each strategy is:

[0063]

[0064] Where M(n) is the comprehensive timing metric value;

[0065] M Sch (n) represents the timing metric value of the nth data position of the data signal obtained using the Schmidl strategy;

[0066] M Minn (n) represents the timing metric value of the nth data position of the data signal obtained using the Minn strategy;

[0067] M Sym (n+N / 2) represents the timing measurement value of the n+N / 2th data position of the data signal obtained by adopting the symmetrical structure training sequence synchronization strategy.

[0068] Preferably, the formula for performing symbol synchronization based on the comprehensive timing metric value is:

[0069]

[0070] Where M(n) is the comprehensive timing metric value; is the symbol synchronization position, and argmax(·) means finding the maximum value.

[0071] Preferably, the frequency offset estimation is performed based on the correlation value of the symbol synchronization position, and the formula is as follows:

[0072]

[0073] in, is the frequency deviation estimate;

[0074] Indicates the symbol synchronization position obtained using the Schmidl strategy The relevant value of

[0075] Indicates the symbol synchronization position obtained using the Minn strategy The relevant value of

[0076] represents the first The relevant value of each data position;

[0077] angle(·) represents the corresponding angle.

[0078] The second aspect of the present invention proposes a symbol synchronization and frequency offset estimation system in an OFDM system based on a CHU sequence, comprising:

[0079] A sequence generation module is used for the transmitter to obtain a conjugate symmetric original sequence U based on the CHU sequence concatenation, and to generate a conjugate antisymmetric training sequence based on the conjugate symmetric original sequence U;

[0080] A data transmission module, used for adding the training sequence into a data frame of a signal in an OFDM system, wherein the data signal including the training sequence is transmitted in a data transmission channel;

[0081] A calculation module, used for receiving a data signal including a training sequence at a receiving end, respectively using a Schmidl strategy, a Minn strategy and a symmetric structure training sequence synchronization strategy to calculate correlation values ​​of different data positions of the data signal based on the training sequence structure, calculating corresponding timing metric values ​​based on the correlation values, and calculating a comprehensive timing metric value based on the timing metric values ​​calculated by each strategy;

[0082] The synchronization module is used for performing symbol synchronization of the OFDM system based on the comprehensive timing measurement value at the receiving end, and performing frequency offset estimation based on the correlation value of the symbol synchronization position.

[0083] Compared with the prior art, the beneficial effects of the present invention include at least:

[0084] 1. The training sequence structure used for symbol synchronization and frequency offset estimation in OFDM systems based on CHU sequences:

[0085] The present invention obtains a conjugate symmetric original sequence U based on CHU sequence splicing, generates a conjugate antisymmetric training sequence based on the conjugate symmetric original sequence U, and reasonably sets the training sequence structure by combining the original sequence U and the conjugate sequence and the opposite sequence of the original sequence. When the sliding window deviates from the correct position, the correlation decreases rapidly.

[0086] The training sequence structure of the present invention complies with the characteristics of the Schmidl repetition structure and the Minn cyclic repetition structure and the characteristics of the symmetric structure training sequence. The proposed training sequence structure and the corresponding timing measurement strategy are combined to eliminate the interference of the secondary peak at the ±N / 4 position. When the symbol synchronization position of the OFDM system obtained by symbol synchronization is within the protection interval, the peak platform effect is eliminated. When the signal-to-noise ratio is reduced, the training sequence can maintain the autocorrelation performance, and the main peak peak is less affected by the reduction of the signal-to-noise ratio.

[0087] The present invention arranges the training sequence, and after the receiving end receives the training sequence, the receiving end converts the training sequence through an operator to achieve a training sequence structure feature that conforms to multiple classic algorithms. The features of the training sequence can be fully extracted to improve the accuracy of symbol synchronization, especially the synchronization accuracy in low OSNR scenarios.

[0088] 2. The training sequence based on the CHU sequence realizes symbol synchronization, and the corresponding correlation value and timing metric calculation of the cross-correlation and auto-correlation combination:

[0089] The present invention respectively adopts the Schmidl strategy, the Minn strategy and the symmetric structure training sequence synchronization strategy to calculate the correlation values ​​of different data positions of the signal for the proposed training sequence, calculates the corresponding timing measurement value based on the correlation value, and calculates the comprehensive timing measurement value based on the timing measurement value calculated by each strategy, thereby combining the advantages of multiple strategies and fully extracting the characteristics of the training sequence. Compared with strategies based on repetitive structure training sequences such as the Schmidl algorithm and the Minn algorithm, the present invention has a higher synchronization position judgment accuracy at high OSNR. Compared with synchronization strategies based on symmetric structure training sequences, the present invention has a higher synchronization position judgment accuracy at low OSNR.

[0090] The present invention inherits the advantages of the cross-correlation and autocorrelation symbol synchronization methods, has high accuracy in symbol synchronization and frequency deviation estimation, and is conducive to hardware implementation. The present invention inherits the high main peak value and strong anti-interference ability of the Minn algorithm and the Schmidl strategy at low OSNR, and inherits the advantages of the synchronization strategy based on the symmetrical structure training sequence at high OSNR symbol synchronization accuracy, low secondary peak value, and high main peak value; the present invention solves the problem that errors may occur in the peak platform and frequency deviation of the Schmidl algorithm; solves the problem that errors occur in the secondary peak interference of the Minn algorithm and frequency deviation estimation; solves the problem that the synchronization strategy based on the symmetrical structure training sequence has low symbol frequency synchronization accuracy at low OSNR. Compared with the method based on a single timing measurement calculation strategy to achieve symbol frequency synchronization, the present invention can more fully extract the characteristics of the training sequence, and has higher symbol synchronization and frequency deviation estimation accuracy, so the application of the present invention can improve the comprehensive performance of the communication system.

[0091] 3. Frequency offset estimation based on the training sequence of CHU sequence:

[0092] Based on the phases of multiple correlation values ​​calculated using different correlation value calculation strategies for the training sequence, multiple frequency offset estimates are obtained, and the average value of the multiple frequency offset estimates is calculated to obtain a comprehensive frequency offset estimation value, thereby improving the frequency offset estimation accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] Figure 1 is the OFDM frame structure used for symbol synchronization;

[0094] Figure 2 is the training sequence structure of the Schmidl algorithm;

[0095] Figure 3 is the training sequence structure of the Minn algorithm;

[0096] Figure 4 is the training sequence structure of the Park algorithm;

[0097] Figure 5 The training sequence structure proposed by the present invention;

[0098] Figure 6 The present invention is a flowchart of the implementation of the method for symbol synchronization and frequency offset estimation in an OFDM system based on a CHU sequence. DETAILED DESCRIPTION

[0099] In order to more clearly illustrate the purpose, technical solutions and advantages of the present invention, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. The embodiments described in this application are embodiments of a part of the present invention, rather than all embodiments. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the protection scope of the present invention.

[0100] Embodiment 1 of the present invention provides a method for jointly synchronizing symbols and frequencies in an OFDM system based on a CHU sequence, which specifically includes the following steps:

[0101] Step 1: The transmitter obtains a conjugate symmetric original sequence U based on the CHU sequence concatenation, and generates a conjugate antisymmetric training sequence based on the conjugate symmetric original sequence U;

[0102] In specific implementation, the single-carrier communication system is sensitive to symbol synchronization errors. The OFDM system is sensitive to symbol synchronization errors and frequency offsets. When there is a synchronization error or frequency offset, inter-carrier interference (ICI) or inter-symbol interference (ISI) will occur, resulting in reduced system performance. Therefore, it is very important for communication systems to maintain accurate synchronization. The training sequence used in the method proposed in the present invention is composed of the conjugate sequence and the opposite sequence of the original number sequence U and U, and a conjugate symmetric sequence composed of the CHU sequence is used as the original number sequence U to perform symbol synchronization and frequency offset estimation, such as Figure 5 shown.

[0103] The training sequence proposed by the present invention consists of 8 parts, U represents a sequence with a length of N / 8, which is generated by splicing N / 16 CHU sequences. -U is the opposite number sequence of U, U* is the conjugate sequence of U, and -U* is the opposite number sequence of the conjugate sequence of U, that is, the conjugate opposite number sequence. The training sequence structure of the method proposed by the present invention is a conjugate antisymmetric sequence.

[0104] Figure 5 The first N / 8 point sequence and the third N / 8 point sequence in the left half of the structure are symmetric sequences; the first N / 8 point sequence and the third N / 8 point sequence in the right half are antisymmetric sequences; the first N / 8 point sequence and the second N / 8 point sequence in the right half are opposite sequences. By using the original sequence, conjugate sequence and opposite sequence of sequence U, the training sequence structure is reasonably set. When the sliding window deviates from the correct position, the correlation value of the training sequence drops rapidly. When the signal-to-noise ratio of the optical fiber channel decreases, the training sequence can maintain the correlation value, and the peak value of the main peak is less affected by the decrease in the signal-to-noise ratio.

[0105] Furthermore, the present invention uses the CHU sequence to conjugate the original symmetric sequence U as the original sequence of the training sequence structure used in the method proposed by the present invention. The expression of the U sequence is:

[0106] U=[CHU conj(CHU')] (3)

[0107] Among them, the CHU sequence is an N / 16 point sequence, ' represents the symmetric operator of the sequence, and the function expression for generating the CHU sequence is:

[0108]

[0109] Wherein, L is the length of the CHU sequence; r is an integer coprime with L; l is a sequence value variable, l = 0, 1…, L-1.

[0110] Step 2: adding the training sequence to a data frame of a signal in an OFDM system, and transmitting the data signal including the training sequence in a data transmission channel;

[0111] Step 3: The receiving end receives a data signal including a training sequence, and respectively uses the Schmidl strategy, the Minn strategy, and the symmetric structure training sequence synchronization strategy to calculate correlation values ​​of different data positions of the data signal based on the training sequence structure, calculates corresponding timing metric values ​​based on the correlation values, and calculates a comprehensive timing metric value based on the timing metric values ​​calculated by the above strategies;

[0112] Further preferably, the present invention uses a symbol synchronization method and a timing metric calculation method combining cross-correlation and auto-correlation, which are specifically as follows:

[0113] The symmetric structure training sequence synchronization strategy uses formula (5) to calculate the correlation value P based on the antisymmetric structure of [-U,-U,-U*,U*,-U*,U*,U,U] for the sequence in the training sequence position corresponding to the nth data position of the received data signal. Sym (n), and calculate the timing metric value M by combining formulas (6) and (7) Sym (n), as follows:

[0114]

[0115]

[0116]

[0117] Wherein, Uc(n) is the nth data of the local sequence [U,U,U*,U*,U*,U*,U,U] pre-shared between the transmitter and the receiver; Uc(m), Uc(N / 2+1-m), and Uc(m+N / 2) are the mth, N / 2+1-m, and m+N / 2th data of the local sequence;

[0118] r(n) represents the nth signal data received, r(n+1-m) and r(n+m) represent the n+1-m and n+mth data received by the receiving end respectively.

[0119] P Sym (n) represents the correlation value of the nth data position of the signal obtained by adopting the symmetric structure training sequence synchronization strategy, which is obtained by the correlation value of the product of the left half sequence and the local sequence and the product of the right half sequence and the local sequence.

[0120] R Sym (n) represents half of the energy value of the sequence in the training sequence position corresponding to the nth data position of the signal obtained by using the symmetric structure training sequence synchronization strategy, which is used for P Sym (n) coefficient for normalization;

[0121] M Sym (n) represents the timing metric value of the nth data position of the signal obtained by using the symmetrical structure training sequence synchronization strategy. The symbol position obtained by the timing metric value of the symmetrical structure training sequence synchronization strategy is the middle position of the training sequence.

[0122] Based on the above symmetric structure training sequence synchronization strategy, the correlation value P of the -N / 4 position of the symbol synchronization position Sym In (n), since the first and second parts of the training sequence are opposite sequences, and the third and fourth parts are the same, the correlation value calculation method based on formula (5) is obtained The value of is lower.

[0123] If the cyclic prefix is ​​filled with signal data, based on the correlation value calculation method of formula (5), the signal data and the training sequence r(n+m), N / 4+1<=m<=N / 2 have no correlation, that is, The value of is lower, and the correlation value and timing measurement value of the -N / 4 position of the symbol synchronization position are lower.

[0124] Based on the symmetric structure training sequence synchronization strategy, the correlation value P of the +N / 4 position of the symbol synchronization position Sym In (n), since the 5th and 6th parts of the training sequence are opposite sequences, and the 7th and 8th parts are the same, the correlation value calculation method based on formula (5) is obtained The value of is lower.

[0125] In addition, the training sequence r(n+m), N / 4+1<=m<=N / 2 has no correlation with the signal data. The correlation value calculation method based on formula (5) is obtained

[0126] The value of is low, and the correlation value and timing metric value of the +N / 4 position of the symbol synchronization position are low. Due to the combination of the training sequence and the timing metric method, the correlation value of the ±N / 4 position of the symbol synchronization position is low, which solves the problem of secondary peak interference at the ±N / 4 position of the symmetrical structure training sequence synchronization strategy. In addition, due to the good cross-correlation property of the CHU sequence, the symbol synchronization and the timing metric value of the position other than the ±N / 4 position of the symbol synchronization are low.

[0127] Further preferably, when the Schmidl strategy or the Minn strategy is used to calculate the timing metric value of the received data signal, the sequence is converted into a training sequence structure of [U,U,U*,U*,U,U,U*,U*] or [U,U,-U*,-U*,U,U,-U*,-U*] by adding an opposite number or a conjugate operator to a specific interval of the training sequence position corresponding to the nth data position of the received signal, which conforms to the training sequence structure of the Schmidl algorithm or the Minn algorithm. Therefore, the Schmidl and Minn timing metric calculation methods can be applied, and based on the Schmidl and Minn training sequence structures, the timing metric calculation method combining cross-correlation and autocorrelation is applied, as follows:

[0128] The Schmidt strategy uses formula (8) to add a negative number and a conjugate operator to the training sequence position corresponding to the nth data position of the received data signal, and calculates the correlation value P of the sequence obtained by adding the negative number and the conjugate operator based on the repetitive structure of [U,U,U*,U*,U,U,U*,U*] Sch (n), and calculate the timing metric value M by combining formulas (9) and (10) Sch (n):

[0129]

[0130]

[0131]

[0132] The Minn strategy uses formula (11) to add the opposite number and the conjugate operator to the training sequence position corresponding to the nth data position of the received data signal, and calculates the correlation value P based on the cyclic repetition structure of [U,U,-U*,-U*,U,U,-U*,-U*] by adding the opposite number and the conjugate operator to the sequence obtained. Minn (n), and calculate the timing metric value M by combining formulas (12) and (13) Minn (n):

[0133]

[0134]

[0135]

[0136] The method proposed by the present invention calculates the timing metric based on the characteristics that the training sequence conforms to Schmidl, Minn and symmetric structure training sequence structures.

[0137] P Sch (n) or P Minn (n) represents the correlation value of the nth signal data position calculated by the timing measurement strategy of the Schmidl algorithm or the Minn algorithm.

[0138] R Sch (n) and R Minn (n) is half of the sequence energy value in the training sequence position corresponding to the nth data position, is the coefficient used for normalization processing, and the correlation value P Sch (n) or P Minn (n) Through R Sch (n) or R Minn The symbol position obtained by the timing metric values ​​of the Schmidl strategy and the Minn strategy is the starting position of the training sequence.

[0139] M Sch (n), M Minn (n), M Sym (n) are the timing metrics of the nth data position obtained by the Schmidl strategy, Minn strategy, and symmetric structure training sequence strategy, respectively.

[0140] For the above Minn strategy, the correlation value P at the -N / 4 position of the symbol synchronization position is Minn (n), since the product of the 3rd and 5th parts of the training sequence is equal to the product of the 4th and 6th parts, the correlation value calculation method based on formula (11) is obtained The value of is low. If the cyclic prefix is ​​filled with signal data, the signal data and the training sequence have no correlation, and the correlation value calculation method based on formula (11) is obtained The value of is lower, and the correlation value and timing measurement value of the -N / 4 position of the symbol synchronization position are lower.

[0141] For the Minn strategy, the correlation value P of the symbol synchronization position +N / 4 Minn In (n), since the product of the 3rd and 5th parts of the training sequence is equal to the product of the 4th and 6th parts, the correlation value calculation method based on formula (11) is obtained The value of is low. In addition, there is no correlation between the signal data and the training sequence. The correlation value calculation method based on formula (11) is obtained The value of is low, and the correlation value and timing metric value of the +N / 4 position of the symbol synchronization position are low. Due to the combination of the training sequence and the timing metric method, the correlation value of the ±N / 4 position of the symbol synchronization position is low, which solves the problem of secondary peak interference at the ±N / 4 position of the symbol synchronization position of the Minn strategy. In addition, due to the good cross-correlation property of the CHU sequence, the symbol synchronization and the timing metric values ​​of the positions other than the ±N / 4 position of the symbol synchronization are low.

[0142] In the classic Schmidl strategy, a part of the end of the training sequence is often used as the prefix of the data frame. For the training sequence structure and Schmidl strategy of this patent, when the OFDM symbol position obtained by symbol synchronization is within the protection interval, the obtained training sequence data is not a cyclic shift of the training sequence, eliminating the peak platform effect. In addition, the timing metric values ​​of Schmidl, Minn and symmetric structure training sequence synchronization strategies that deviate from the correct symbol synchronization position are low, and the secondary peaks of the above strategies appear at different positions, so the mean amplitude of the timing metric of the above strategies that deviate from the correct symbol synchronization position is low. Near the correct symbol synchronization position, Schmidl, Minn and symmetric structure training sequence synchronization strategies have timing metrics of different shapes and higher values ​​of timing metrics, so the mean of the timing metric of the above strategies is higher at the correct symbol synchronization position and there is no platform problem.

[0143] Further preferably, in order to Sch (n), M Minn (n), M Sym (n) linear addition, it is necessary to consider that the symbol synchronization position obtained by different strategies is offset from the starting position of the training sequence. The symbol synchronization position of the Schmidl and Minn strategies is the starting position of the training sequence, and the symbol synchronization position obtained by the symbol synchronization strategy based on the symmetric structure training sequence is the middle position of the training sequence. Therefore, in order to linearly add the timing metrics of different strategies, when using the timing metric M of the symmetric structure training sequence Sym (n), the timing metric of n+N / 2 data positions needs to be used, that is, M Sym (n+N / 2), align the synchronization position.

[0144] The comprehensive timing metric M(n) is given by M Sch (n), M Minn (n), M Sym The arithmetic mean of (n+N / 2) is obtained as follows:

[0145]

[0146] Where M(n) is the comprehensive timing metric value;

[0147] MSch (n) represents the timing metric value of the nth data position in the data signal obtained using the Schmidl strategy;

[0148] M Minn (n) represents the timing metric value of the nth data position in the data signal obtained using the Minn strategy;

[0149] M Sym (n+N / 2) represents the timing measurement value of the n+N / 2th data position in the data signal obtained by adopting the symmetrical structure training sequence synchronization strategy.

[0150] Step 4: The receiver performs symbol synchronization of the OFDM system based on the comprehensive timing metric value and estimates the frequency offset based on the correlation value of the symbol synchronization position, as follows:

[0151]

[0152] Where M(n) is the comprehensive timing metric value; is the symbol synchronization position, and argmax(·) means finding the maximum value.

[0153] Further preferably, the method further comprises: performing frequency offset estimation based on a correlation value of a symbol synchronization position:

[0154] The training sequence structure proposed by the present invention meets the characteristics of Schmidl, Minn and symmetric structure training sequence structures. The estimated value that can be calculated using the corresponding frequency offset estimation strategy and The frequency offset estimation obtained by different strategies is defined as follows:

[0155]

[0156]

[0157]

[0158]

[0159]

[0160] in, is the frequency deviation estimate;

[0161] Indicates the symbol synchronization position obtained using the Schmidl strategy The relevant value of

[0162] Indicates the symbol synchronization position obtained using the Minn strategy The relevant value of

[0163] represents the correlation value of the symmetric structure training sequence, that is, the first The relevant value of each data location. and The range is [-ππ], and the fractional part of the frequency offset estimation method proposed in the present invention The range is [-0.5 0.5].

[0164] In summary, if Figure 6 As shown, the implementation steps of the method proposed by the present invention for symbol synchronization and frequency offset estimation are as follows:

[0165] 1) Use formula (4) to generate an N / 16-point CHU sequence.

[0166] 2) Use formula (3) to concatenate the CHU sequence into the conjugate symmetric original sequence U.

[0167] 3) Based on the sequence U, Figure 5 The training sequence shown is composed of 8 parts, U represents a sequence of length N / 8. -U is the opposite sequence of U, U* is the conjugate sequence of U, and -U* is the opposite conjugate sequence of U. The structure of the above training sequence is a conjugate antisymmetric sequence.

[0168] 4) The sending end adds the training sequence to the data frame, and the data frame is transmitted in the data transmission channel.

[0169] 5) The receiving end receives the data frame including the training sequence.

[0170] When calculating the timing metric value of the received data signal, the sequence is converted into [U,U,U*,U*,U,U,U*,U*] or [U,U,-U*,-U*,U,U,-U*,-U*] by adding an opposite number or a conjugate operator to a specific interval of the expected received training sequence [-U,-U,-U*,U*,-U*,U*,U,U], which conforms to the training sequence structure of the Schmidl algorithm or the Minn algorithm. Therefore, the method of the present invention can calculate the correlation value and the timing metric by using the timing metric method combining cross-correlation and auto-correlation based on the Schmidl strategy, the Minn strategy and the synchronization strategy of the symmetrical structure training sequence.

[0171] 6) The receiving end uses a timing measurement method combining cross-correlation and autocorrelation based on the training sequence structure of the Schmidl strategy, and calculates the correlation value and timing measurement value of different data positions of the data signal based on formulas (8)-(10).

[0172] 7) The receiving end uses the Minn strategy-based training sequence structure to use a timing measurement method that combines cross-correlation and autocorrelation to calculate the correlation values ​​and timing measurement values ​​of different data positions of the data signal based on formulas (11)-(13).

[0173] 8) The receiving end uses a timing measurement method combining cross-correlation and autocorrelation based on the synchronization strategy of the symmetrical structure training sequence, and calculates the correlation value and timing measurement value of different data positions of the data signal based on formulas (5)-(7).

[0174] 9) The receiving end calculates the comprehensive timing measurement value of the method of the present invention based on formula (14).

[0175] 10) The receiving end performs symbol synchronization based on the timing measurement of the signal, and the symbol synchronization position From formula (15), we can see that the maximum value of the timing metric value is the position of the training sequence. Based on the position of the training sequence, we can get the synchronization position of the data frame and achieve symbol synchronization of the data signal.

[0176] 11) The receiving end estimates the frequency offset based on the correlation value of the symbol synchronization position using formula (20).

[0177] Embodiment 2 of the present invention provides a symbol synchronization and frequency offset estimation system in an OFDM system based on a CHU sequence, including:

[0178] A sequence generation module is used for the transmitter to obtain a conjugate symmetric original sequence U based on the CHU sequence concatenation, and to generate a conjugate antisymmetric training sequence based on the conjugate symmetric original sequence U;

[0179] A data transmission module, used for adding the training sequence into a data frame of a signal in an OFDM system, wherein the data signal including the training sequence is transmitted in a data transmission channel;

[0180] A calculation module, used for receiving a data signal including a training sequence at a receiving end, respectively using a Schmidl strategy, a Minn strategy and a symmetric structure training sequence synchronization strategy to calculate correlation values ​​of different data positions of the data signal based on the training sequence structure, calculating corresponding timing metric values ​​based on the correlation values, and calculating a comprehensive timing metric value based on the timing metric values ​​calculated by each strategy;

[0181] The synchronization module is used for performing symbol synchronization of the OFDM system based on the comprehensive timing measurement value at the receiving end, and performing frequency offset estimation based on the correlation value of the symbol synchronization position.

[0182] Embodiment 3 of the present invention provides a terminal, including a processor and a storage medium; the storage medium is used to store instructions; the processor operates based on the instructions to execute the steps of the method.

[0183] Embodiment 4 of the present invention provides a computer-readable storage medium on which a computer program is stored. When the stored program is executed by a processor, the steps of the method are implemented.

[0184] Compared with the prior art, the beneficial effects of the present invention include at least:

[0185] 1. The training sequence structure used for symbol synchronization and frequency offset estimation in OFDM systems based on CHU sequences:

[0186] The present invention obtains a conjugate symmetric original sequence U based on CHU sequence splicing, generates a conjugate antisymmetric training sequence based on the conjugate symmetric original sequence U, and reasonably sets the training sequence structure by combining the original sequence U and the conjugate sequence and the opposite sequence of the original sequence, so that the correlation values ​​of different parts of the training sequence are rapidly reduced when the sliding window deviates from the correct position;

[0187] The training sequence structure of the present invention complies with the characteristics of the Schmidl repetition structure and the Minn cyclic repetition structure and the characteristics of the symmetrical structure training sequence. The proposed training sequence structure and the corresponding timing measurement strategy are combined to eliminate the interference of the secondary peak at the ±N / 4 position. When the OFDM symbol synchronization position obtained by symbol synchronization is within the protection interval, the peak platform effect is eliminated. When the signal-to-noise ratio is reduced, the training sequence can maintain the autocorrelation performance, and the main peak peak is less affected by the reduction of the signal-to-noise ratio.

[0188] After receiving the training sequence, the receiving end of the present invention transforms the training sequence through an operator to achieve a training sequence structure feature that conforms to multiple classical algorithms, which can fully extract the features of the training sequence and improve the accuracy of symbol synchronization, especially the synchronization accuracy in low OSNR scenarios.

[0189] 2. The training sequence based on the CHU sequence realizes symbol synchronization, and the corresponding correlation value and timing metric calculation of the cross-correlation and auto-correlation combination:

[0190] The present invention adopts the Schmidl strategy, the Minn strategy and the symmetric structure training sequence synchronization strategy to calculate the correlation values ​​of different data positions of the signal for the proposed training sequence, calculates the corresponding timing metric value based on the correlation value, calculates the comprehensive timing metric value based on the timing metric value calculated by each strategy, and comprehensively extracts the characteristics of the training sequence by combining the advantages of multiple strategies. Compared with the strategies based on the cyclic repetitive structure training sequence such as Schmidl and Minn, it has a higher synchronization accuracy rate at high OSNR, and compared with the strategy based on the symmetric structure training sequence, it has a higher synchronization accuracy rate at low OSNR.

[0191] The present invention inherits the advantages of the cross-correlation and auto-correlation symbol synchronization methods, has high symbol synchronization accuracy, and is conducive to hardware implementation.

[0192] The present invention inherits the advantages of high main peak value and strong anti-interference ability of Minn and Schmidl strategies at low OSNR, and inherits the advantages of high symbol synchronization accuracy, low secondary peak value and high main peak value of synchronization strategy based on symmetrical structure training sequence at high OSNR; the present invention solves the problem of large error in peak platform and frequency deviation estimation of Schmidl strategy; solves the problem of error in secondary peak interference with main peak and frequency deviation estimation of Minn strategy; solves the problem of low symbol frequency synchronization accuracy of synchronization strategy based on symmetrical structure training sequence at low OSNR. Compared with the symbol synchronization strategy based on single timing metric calculation method, the present invention can more fully extract the characteristics of training sequence and has higher symbol synchronization accuracy, so the application of the present invention can improve the comprehensive performance of communication system.

[0193] 3. Frequency offset estimation based on the training sequence of CHU sequence:

[0194] Multiple frequency offset estimates are obtained based on multiple correlation values ​​calculated using different correlation value calculation strategies for the training sequence, and the average value of the multiple frequency offset estimates is calculated to obtain a comprehensive frequency offset estimate, thereby improving the accuracy of the frequency offset estimation.

[0195] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention shall fall within the scope of protection of the claims of the present invention.

Claims

1. A method for symbol synchronization and frequency offset estimation in an OFDM system based on a CHU sequence, characterized in that: include: The transmitter obtains a conjugate symmetric original sequence U based on the CHU sequence concatenation, and generates a conjugate antisymmetric training sequence based on the conjugate symmetric original sequence U; Adding the training sequence to a data frame of a signal in an OFDM system, wherein the data signal including the training sequence is transmitted in a data transmission channel; The receiving end receives a data signal including a training sequence, and respectively uses a Schmidl strategy, a Minn strategy, and a symmetric structure training sequence synchronization strategy to calculate correlation values ​​of different data positions of the data signal based on the training sequence structure, calculates corresponding timing metric values ​​based on the correlation values, and calculates a comprehensive timing metric value based on the timing metric values ​​calculated by each strategy; The receiving end performs symbol synchronization of the OFDM system based on the comprehensive timing measurement value, and performs frequency offset estimation based on the correlation value of the symbol synchronization position.

2. The method for symbol synchronization and frequency offset estimation in an OFDM system based on a CHU sequence according to claim 1, characterized in that: The expression of the conjugated symmetric original sequence U obtained by splicing the CHU sequence is: U=[CHU conj(CHU')] (3) Wherein, U represents the conjugate symmetric original sequence of length N / 8, CHU represents the CHU sequence of N / 16 points, N represents the training sequence length, CHU' represents the symmetric sequence of the CHU sequence, and conj represents the conjugate of the sequence; The lth variable in the CHU sequence is: Wherein, L is the length of the CHU sequence; r is an integer coprime with L; l = 0, 1…, L-1.

3. The method for joint synchronization of OFDM symbols and frequencies based on CHU sequence according to claim 2, characterized in that: Generate a conjugate antisymmetric training sequence based on the conjugate symmetric original sequence U: [-U,-U,-U*,U*,-U*,U*,U,U]; Among them, -U is the opposite number sequence of U, U* is the conjugate sequence of U, and -U* is the conjugate opposite number sequence of U.

4. The method for symbol synchronization and frequency offset estimation in an OFDM system based on a CHU sequence according to claim 3, characterized in that: The Schmidt strategy uses formula (8) to add a negative number and a conjugate operator to the training sequence position corresponding to the nth data position of the received data signal, and calculates the correlation value P of the sequence obtained by adding the negative number and the conjugate operator based on the repetitive structure of [U,U,U*,U*,U,U,U*,U*] Sch (n), and calculate the timing metric value M by combining formulas (9) and (10) Sch (n): in, P Sch (n) represents the correlation value of the nth data position in the data signal obtained using the Schmidl strategy; Uc(m),U c (m+N / 2) is the m-th and m+N / 2-th data of the pre-shared local sequence [U,U,U*,U*,U*,U*,U,U]; r(n+m) and r(n+m+N / 2) respectively represent the n+m and n+m+N / 2th data in the training sequence position corresponding to the nth data position in the data signal received by the receiving end; R Sch (n) is used for P Sch (n) coefficient for normalization; M Sch (n) represents the timing metric value of the nth data position in the data signal obtained using the Schmidl strategy; m is the relative position variable relative to n.

5. The method for symbol synchronization and frequency offset estimation in an OFDM system based on CHU sequence according to claim 3, characterized in that: The Minn strategy uses formula (11) to add the opposite number and the conjugate operator to the training sequence position corresponding to the nth data position of the received data signal, and calculates the correlation value P based on the cyclic repetition structure of [U,U,-U*,-U*,U,U,-U*,-U*] by adding the opposite number and the conjugate operator to the sequence obtained. Minn (n), and calculate the timing metric value M by combining formulas (12) and (13) Minn (n): Among them, P Minn (n) represents the correlation value of the nth data position in the data signal obtained by using the Minn strategy; Uc(m), Uc(m+(k-1)N / 8), Uc(m+N / 4+(k-1)N / 8), Uc(m+N / 2+(k-1)N / 8), and Uc(m+3N / 4+(k-1)N / 8) are the m, m+(k-1)N / 8, m+N / 4+(k-1)N / 8, m+N / 2+(k-1)N / 8, and m+3N / 4+(k-1)N / 8 data of the pre-shared local sequence [U,U,U*,U*,U*,U*,U,U]; r(n+m), r(n+m+(k-1)N / 8), r(n+m+N / 4+(k-1)N / 8), r(n+m+N / 2+(k-1)N / 8), and r(n+m+3N / 4+(k-1)N / 8) respectively represent the n+m, n+m+(k-1)N / 8, n+m+N / 4+(k-1)N / 8, and n+m+3N / 4+(k-1)N / 8 data in the training sequence position corresponding to the nth data position in the data signal received by the receiving end; R Minn (n) is used for P Minn (n) coefficient for normalization; M Minn (n) represents the timing metric value of the nth data position in the data signal obtained using the Minn strategy; k is the control variable and m is the relative position variable relative to n.

6. The method for symbol synchronization and frequency offset estimation in an OFDM system based on a CHU sequence according to claim 3, characterized in that: The symmetric structure training sequence synchronization strategy uses formula (5) to calculate the correlation value P of the sequence in the training sequence position corresponding to the nth data position of the received data signal based on the antisymmetric structure of [-U,-U,-U*,U*,-U*,U*,U,U] Sym (n), and calculate the timing metric value MS by combining formulas (6) and (7) ym (n): Among them, P Sym (n) represents the correlation value of the nth data position in the data signal obtained by using the symmetric structure training sequence synchronization strategy; Uc(m), Uc(N / 2+1-m), and Uc(m+N / 2) are the mth, N / 2+1-mth, and m+N / 2th data of the local sequence [U,U,U*,U*,U*,U*,U] pre-shared between the transmitter and the receiver; r(n+1-m) and r(n+m) respectively represent the n+1-mth and n+mth data in the training sequence position corresponding to the nth data position in the data signal received by the receiving end; R Sym (n) is used for P Sym (n) coefficient for normalization; M Sym (n) represents the timing metric value of the nth data position in the data signal obtained by using the symmetric structure training sequence synchronization strategy; k is the control variable and m is the relative position variable relative to n.

7. The method for symbol synchronization and frequency offset estimation in an OFDM system based on CHU sequence according to claim 1, characterized in that: The formula for calculating the comprehensive timing metric value based on the timing metric values ​​calculated by each strategy is: Where, M(n) is the comprehensive timing metric value; M Sch (n) represents the timing metric value of the nth data position of the data signal obtained using the Schmidl strategy; M Minn (n) represents the timing metric value of the nth data position of the data signal obtained using the Minn strategy; M Sym (n+N / 2) represents the timing measurement value of the n+N / 2th data position of the data signal obtained by adopting the symmetrical structure training sequence synchronization strategy.

8. The method for symbol synchronization and frequency offset estimation in an OFDM system based on CHU sequence according to claim 1, characterized in that: The formula for symbol synchronization based on the comprehensive timing metric value is: Where, M(n) is the comprehensive timing metric value; is the symbol synchronization position, and argmax(·) means finding the maximum value.

9. The method for symbol synchronization and frequency offset estimation in an OFDM system based on CHU sequence according to claim 1, characterized in that: The frequency offset estimation is performed based on the correlation value of the symbol synchronization position, and the formula is as follows: in, is the frequency deviation estimate; Indicates the symbol synchronization position obtained using the Schmidl strategy The relevant value of Indicates the symbol synchronization position obtained using the Minn strategy The relevant value of represents the first The relevant value of each data position; angle(·) represents the corresponding angle.

10. A symbol synchronization and frequency offset estimation system in an OFDM system based on a CHU sequence, using the method according to any one of claims 1 to 9, characterized in that: The system comprises: A sequence generation module is used for the transmitter to obtain a conjugate symmetric original sequence U based on the CHU sequence concatenation, and to generate a conjugate antisymmetric training sequence based on the conjugate symmetric original sequence U; A data transmission module, used for adding the training sequence into a data frame of a signal in an OFDM system, wherein the data signal including the training sequence is transmitted in a data transmission channel; A calculation module, used for receiving a data signal including a training sequence at a receiving end, respectively using a Schmidl strategy, a Minn strategy and a symmetric structure training sequence synchronization strategy to calculate correlation values ​​of different data positions of the data signal based on the training sequence structure, calculating corresponding timing metric values ​​based on the correlation values, and calculating a comprehensive timing metric value based on the timing metric values ​​calculated by each strategy; The synchronization module is used for performing symbol synchronization of the OFDM system based on the comprehensive timing measurement value at the receiving end, and performing frequency offset estimation based on the correlation value of the symbol synchronization position.

Images