CN111193681A

CN111193681A - Time synchronization method based on high-precision data transmission

Info

Publication number: CN111193681A
Application number: CN201911305353.1A
Authority: CN
Inventors: 李小红; 张吉林; 沈亮; 李文军; 王维
Original assignee: Ksw Technologies Co ltd
Current assignee: Ksw Technologies Co ltd
Priority date: 2019-02-28
Filing date: 2019-02-28
Publication date: 2020-05-22
Anticipated expiration: 2039-02-28
Also published as: CN110071888A; CN111193681B; CN110071888B

Abstract

The invention discloses a time synchronization method based on high-precision data transmission, which comprises the following steps: step one, generating N_sZC sequences seq with L roots and different indexes_ii=1…N_ｓ(ii) a Step two, the N is_sThe ZC sequences are spliced end to end into LN with the length of_sThe sequence seq of (1); step three, the obtained sequence seq is subjected to subcarrier mapping, then is converted into a time domain, and is subjected to N_sN of different offset_sMultiplying to obtain N_sA plurality of different sequences, which are added to obtain a local sequence S_local(ii) a Step four, N is carried out on the received OFDM signals_sExtracting to obtain a sequence r, and comparing the sequence r with a local sequence S_localPerforming sliding correlation calculation detection to obtain a correlation peak value; and step five, determining the frame header starting position of the signal according to the position of the correlation peak. The invention can realize that the correlation of the synchronous sequence is not influenced by the extraction of the offset point, and can greatly reduce the storage operation resources and the processing required by hardware realizationTime.

Description

Time synchronization method based on high-precision data transmission

Technical Field

The invention relates to the field of communication, in particular to a time synchronization method based on high-precision data transmission.

Background

In a wireless communication system, in order to combat interference caused by multipath transmission, OFDM (orthogonal frequency division multiplexing) or SC-FDE (single carrier frequency domain equalization) is generally used as a communication system. Especially in the telemetry systems of today, the fast development of drones makes the machines more and more powerful in their load-carrying capacity and therefore more demanding in terms of communication rate.

In a burst communication system, data-aided methods are generally used to time synchronize signals. The start position of the OFDM signal is determined by finding the correlation peak position of the local sequence and the received signal. The main bottleneck of the OFDM high data rate transmission system is that a higher transmission data rate requires a larger bandwidth to achieve. The bandwidth directly determines the number of available sub-carrier points for the OFDM system. If the number of available sub-carrier points is large, the storage and operation resources of hardware are greatly required when the time synchronization is used for relevant calculation, and a lot of delay is brought. Therefore, research work on burst transmission links with transmission rates of 300Mbit/s or more is actively being conducted domestically.

For the correlation calculation with larger point number, two ideas are generally adopted, wherein the first is to respectively take symbols of a received signal and a local sequence for correlation; the second is to extract the received signal and the local sequence respectively to an acceptable range for correlation. The method of taking symbols for correlation is greatly affected by system frequency offset, so the method is mainly based on a fast time synchronization scheme of decimation. The main problem with decimation-based time synchronization schemes is the decimation offset. The receiving end does not know the arrival time of the signal, and cannot judge the offset of the current extraction point and the extraction point inside, so that if the received signal and the local sequence are extracted, the correlation of the sequence is damaged. Based on this problem, a new time synchronization scheme is proposed. The correlation of the synchronous sequence of the scheme is not influenced by the extraction offset point, and the storage operation resource and the processing time required by hardware implementation can be greatly reduced.

Disclosure of Invention

The invention aims to overcome the defects of the prior art and provides a time synchronization method based on high-precision data transmission, which comprises the following steps:

step one, generating N_sZC (zadoff-chu) sequences seq with L roots and different indexes_i,i＝1,..,N_s；

Step two, the N is_sIndividual ZC sequences seq_iSpliced end to end into LN with length_sThe sequence seq of (1);

step three, the obtained sequence seq is subjected to subcarrier mapping, then is converted into a time domain, and is subjected to N_sN of different offset_sMultiplying to obtain N_sDifferent sequences are added to obtain a local sequence S_local；

Step four, N is carried out on the received OFDM signals_sExtracting to obtain a sequence r, and comparing the sequence r with a local sequence S_localPerforming sliding correlation calculation detection to obtain a correlation peak value;

and step five, determining the frame header starting position of the signal according to the position of the correlation peak.

Further, the generation formula of the ZC sequence is:

(1) wherein N represents the length of the ZC sequence, R represents the root index, and N and R are relatively prime integers.

Further, the sub-carrier mapping of the spliced sequence seq in step three is that the ZC sequences are arranged on the positive and negative sub-carriers of the OFDM in a way that the 1 st to the N th of the ZC sequences_sThe/2 elements are arranged on the negative subcarrier, Nth _s2+1 to N_sElementThe elements are arranged on positive subcarriers.

Further, performing inverse Fourier transform on the seq sequence obtained in the second step to obtain a sequence seq_time(ii) a Is provided with N_sSet of sequences s_i},i＝1…N_s；s_iIs composed of seq_iLength obtained after zero padding is LN_sThe zero padding mode is shown in formulas (2) to (5):

seq is known from the Fourier transform property_timeComprises the following steps:

further, the local sequence s in step three_localIs a pair seq_timeCarry out N_sAfter double harvest, N_sThe sequence resulting from the addition of sequences generated by offsets is s_localThe calculation method is as follows:

further, the sliding correlation in step four is specifically implemented by taking the point of the current sliding window and L points in the future as a sequence, and using the sequence and the local sequence sl_ocal, performing correlation calculation, wherein the calculation formula is as follows:

where x (i), i ═ 1, …, N, y (i), i ═ 1, …, N represent two complex sequences of length N. conj (A) denotes the conjugate of A, and abs | A | denotes the absolute value of A.

Further, the correlation peak value in the fifth step is a correlation value obtained by sliding the sliding window once, and if the correlation value is greater than the set value and is the maximum value among all the correlation values, the point is the correlation peak position.

Drawings

FIG. 1 is a flow chart of a method for fast time synchronization in high speed data transmission;

FIG. 2 is a schematic view of time synchronization scheme ZC sequence subcarrier mapping;

FIG. 3 is a schematic view of a ZC sequence splicing;

FIG. 4 is a correlation diagram for a new scheme offset of 0;

FIG. 5 is a correlation diagram when the new scheme offset is 1;

FIG. 6 is a correlation diagram for a new scheme offset of 2;

FIG. 7 is a correlation diagram for a new scheme offset of 3;

FIG. 8 is a correlation diagram for a new scheme offset of 4;

FIG. 9 is a correlation diagram for a new scheme offset of 5;

FIG. 10 is a correlation diagram for a new scheme offset of 6;

FIG. 11 is a correlation diagram for a new scheme offset of 7;

FIG. 12 is a correlation diagram for a new scheme offset of 8;

FIG. 13 is a correlation diagram for a new scheme offset of 9;

FIG. 14 is a correlation diagram for a new scheme offset of 10;

FIG. 15 is a correlation diagram for a new scheme offset of 11;

FIG. 16 is a correlation diagram when the offset of the new scheme is 12;

FIG. 17 is a correlation diagram for a new scheme offset of 13;

FIG. 18 is a correlation diagram with a new scheme offset of 14;

FIG. 19 is a correlation diagram for a new scheme offset of 15.

Detailed Description

The technical solutions of the present invention are further described in detail below with reference to the accompanying drawings, but the scope of the present invention is not limited to the following.

Take an OFDM system for 300Mbps high rate data transmission as an example. The system employs a TDMA (time division multiple access) scheme. The subcarrier interval of the system is 15KHz, the number of available subcarriers is 12240, the number of FFT points is 16384, and the length of CP (cyclic prefix) is 1152. And channel coding with a code rate of 16QAM 1/2 is adopted. One physical layer slot is 1ms (millisecond) in length and contains 14 OFDM symbols, wherein the first OFDM symbol is a synchronization header for achieving time synchronization. The multiple of the decimation is 16.

The synchronization header framing procedure is as follows:

according to equation (1), 16 data with length 765 and root index 171113192329313741434753596167 are generated]ZC sequence of seq_i,i＝1,..,16。

seq₁(m+1)＝exp(-j*pi*m*(m+1)/765),m＝0,…,764

seq₂(m+1)＝exp(-j*pi*7*m*(m+1)/765),m＝0,…,764

seq₃(m+1)＝exp(-j*pi*11*m*(m+1)/765),m＝0,…,764

seq₄(m+1)＝exp(-j*pi*13*m*(m+1)/765),m＝0,…,764

seq₅(m+1)＝exp(-j*pi*19*m*(m+1)/765),m＝0,…,764

seq₆(m+1)＝exp(-j*pi*23*m*(m+1)/765),m＝0,…,764

seq₇(m+1)＝exp(-j*pi*29*m*(m+1)/765),m＝0,…,764

seq₈(m+1)＝exp(-j*pi*31*m*(m+1)/765),m＝0,…,764

seq₉(m+1)＝exp(-j*pi*37*m*(m+1)/765),m＝0,…,764

seq₁₀(m+1)＝exp(-j*pi*41*m*(m+1)/765),m＝0,…,764

seq₁₁(m+1)＝exp(-j*pi*43*m*(m+1)/765),m＝0,…,764

seq₁₂(m+1)＝exp(-j*pi*47*m*(m+1)/765),m＝0,…,764

seq₁₃(m+1)＝exp(-j*pi*53*m*(m+1)/765),m＝0,…,764

seq₁₄(m+1)＝exp(-j*pi*59*m*(m+1)/765),m＝0,…,764

seq₁₅(m+1)＝exp(-j*pi*61*m*(m+1)/765),m＝0,…,764

seq₁₆(m+1)＝exp(-j*pi*67*m*(m+1)/765),m＝0,…,764

The 16 ZC sequences seq with the length 765_iI 1., 16 are spliced end to end in the manner shown in fig. 3 to form a sequence seq of length 12240.

The seq subjected to subcarrier mapping is subjected to ifft of 16384 points to obtain a time domain sequence seq with 16384 points in sequence_time. Pair seq_time16 times of extraction of 16 different offsets is carried out, and 16 sequences with length of 1024, seqD, can be obtained_iI is 1, …,16, wherein

The offset 0 corresponds to the sequence:

seqD₁＝[seq_time(1) seq_time(17) seq_time(33)…seq_time(16369)]

the offset is 1 for the sequence:

seqD₂＝[seq_time(2) seq_time(18) seq_time(34)…seq_time(16370)]

the offset 2 corresponds to the sequence:

seqD₃＝[seq_time(3) seq_time(19) seq_time(35)…seq_time(16371)]

the sequence with offset 3 is:

seqD₄＝[seq_time(4) seq_time(20) seq_time(36)…seq_time(16372)]

the offset 4 corresponds to the sequence:

seqD₅＝[seq_time(5) seq_time(21) seq_time(37)…seq_time(16373)]

the offset of 5 corresponds to the sequence:

seqD₆＝[seq_time(6) seq_time(22) seq_time(38)…seq_time(16374)]

the offset is 6 for the sequence:

seqD₇＝[seq_time(7) seq_time(23) seq_time(39)…seq_time(16375)]

the offset of 7 corresponds to the sequence:

seqD₈＝[seq_time(8) seq_time(24) seq_time(40)…seq_time(16376)]

the offset of 8 corresponds to the sequence:

seqD₉＝[seq_time(9) seq_time(25) seq_time(41)…seq_time(16377)]

the offset 9 corresponds to the sequence:

seqD₁₀＝[seq_time(10) seq_time(26) seq_time(42)…seq_time(16378)]

the offset is 10 for the sequence:

seqD₁₁＝[seq_time(11) seq_time(27) seq_time(43)…seq_time(16379)]

the offset of 11 corresponds to the sequence:

seqD₁₂＝[seq_time(12) seq_time(28) seq_time(44)…seq_time(16380)]

the offset is 12 for the sequence:

seqD₁₃＝[seq_time(13) seq_time(29) seq_time(45)…seq_time(16381)]

the offset 13 corresponds to the sequence:

seqD₁₄＝[seq_time(14) seq_time(30) seq_time(46)…seq_time(16382)]

the offset 14 corresponds to the sequence:

seqD₁₅＝[seq_time(15) seq_time(31) seq_time(47)…seq_time(16383)]

the offset 15 corresponds to the sequence:

seqD₁₆＝[seq_time(16) seq_time(32) seq_time(48)…seq_time(16384)]

the 16 sequences are accumulated together and combined into a local sequence

The synchronization process of the receiving end is mainly as follows:

16 times of the received sequence is extracted to obtain the sequences rD and s_localA sliding correlation is performed. The specific implementation manner of the sliding correlation is to take the current time and 1024 points in the future as a sequence, and use the sequence as the local sequence s_localAnd carrying out correlation calculation. I.e. the first point of 1 to 1024 indices and s_localA correlation is performed. Second index of 2 to 1025 points and s_localA correlation is performed. And so on. One correlation value is obtained for each sliding. The correlation value exceeds 10 and is the largest of all correlation values, and this point is denoted as the correlation peak. Since the length of the local sequence is 1024, and decimation is performed_localWhen the obtained receiving sequence is in sliding correlation, the correct time synchronization is realized only if the point corresponding to the peak value is 1024.

Under the environment of-3 dB signal-to-noise ratio, 16 times of extraction is carried out on the receiving sequences, and the offsets are respectively [ 0123456789101112131415 ]]The obtained 16 down-sampled received sequences and local sequences s_localThe results of performing the correlation calculations are shown in fig. 4-19. It can be seen from the figure that the position of each correlation peak is located at the 1024 th point, so that the time synchronization is correctly realized by the down-sampling sequences corresponding to 16 different offsets. Therefore, the correlation between the local sequence and the received sequence in the scheme is not damaged by the offset. Is differentIn the offset scenario, the correlation is still good. Therefore, in the actual implementation process, reasonable extraction multiples can be designed according to the hardware resources of the system. The method can reduce hardware storage resources, operation overhead and signal synchronization time, and can obtain better time synchronization performance.

The foregoing is illustrative of the preferred embodiments of this invention, and it is to be understood that the invention is not limited to the precise form disclosed herein and that various other combinations, modifications, and environments may be resorted to, falling within the scope of the concept as disclosed herein, either as described above or as apparent to those skilled in the relevant art. And that modifications and variations may be effected by those skilled in the art without departing from the spirit and scope of the invention as defined by the appended claims.

Claims

1. A fast time synchronization method in high-speed data transmission is characterized by comprising the following steps:

step one, generating N_sZC sequences seq with L roots and different indexes_i,i＝1,..,N_s；

step three, the obtained sequence seq is subjected to subcarrier mapping, then is converted into a time domain, and is subjected to N_sN of different offset_sMultiplying to obtain N_sA plurality of different sequences, which are added to obtain a local sequence S_local；

Step four, N is carried out on the received OFDM signals_sExtracting to obtain a sequence r, and comparing the sequence r with a local sequence S_localPerforming sliding correlation calculation to obtain a correlation peak value, and determining the position of the correlation peak value;

step five, determining the frame header starting position of the signal according to the position of the correlation peak;

the generation formula of the ZC sequence is as follows:

wherein L represents the length of the ZC sequence, R represents the root index, and N and R are coprime integers;

the sub-carrier mapping of the obtained sequence seq in step three is that the ZC sequence is arranged on the positive and negative sub-carriers of the OFDM in a way that the 1 st to the N th of the ZC sequence_sThe/2 elements are arranged on the negative subcarrier, Nth_s2+1 to N_sThe elements are arranged on positive subcarriers.

2. The method according to claim 1, wherein the sequence seq of step two is subjected to inverse fourier transform to obtain a sequence seq_time(ii) a Is provided with N_sSet of sequences s_i},i＝1…N_s；s_iIs composed of seq_iLength obtained after zero padding is LN_sThe zero padding mode is shown as the formula:

seq_timecomprises the following steps:

3. the method of claim 1, wherein the fast time synchronization method in high speed data transmissionCharacterized by a local sequence s in said step three_localIs a pair seq_timeCarry out N_sAfter double harvest, N_sThe sequence resulting from the addition of sequences generated by offsets is s_localThe calculation method is as follows:

4. the method as claimed in claim 1, wherein the sliding correlation in step four is implemented by taking the current sliding window point and L points after the current sliding window point as a sequence, and the local sequence s_localAnd performing correlation calculation, wherein the formula of the correlation calculation is as follows:

wherein x (i), i ═ 1, …, N, y (i), i ═ 1, …, N represent two complex sequences of length N, respectively; conj (A) denotes the conjugate of A, and abs | A | denotes the absolute value of A.

5. The method according to claim 1, wherein the correlation peak in step five has a value of s and s for each sliding of the sliding window_localAnd performing correlation calculation to obtain a correlation value, wherein if the correlation value is greater than the set value and is the maximum value of all the correlation values, the position of a sliding window of the correlation value is the position of the correlation peak.