CN103346990B - A kind of method of Symbol Timing correction in DVB-T2 system - Google Patents

Abstract

The invention discloses a kind of method of Symbol Timing correction in DVB T2 system, in sign synchronization, use the P1 symbol regenerated in receiver, and reception signal does relevant, finds the original position of OFDM symbol；Related operation is carried out staged operation；Use the time-domain signal of the P2 pilot tone of local generation, do relevant according to the above-mentioned symbol original position docking collection of letters number；The slip result of 5 is done weighted average, GI pattern that then maximizing is corresponding and position.The method of the present invention significantly reduces operand and computing cost, improves mode detection accuracy.

Description

A kind of method of Symbol Timing correction in DVB-T2 system

Technical field

The present invention relates to Symbol Timing and the method for mode detection in DVB-T2 system.

Background technology

In multi-carrier systems, in order to keep the orthogonality of OFDM symbol, and OFDM is reduced to greatest extent (ISI) is interfered, it will usually between two OFDM, insert protection interval (GI), come between symbol Reach above-mentioned purpose.I.e. afterbody from previous OFDM symbol takes a certain proportion of part and copies to currently The original position of OFDM symbol.

In multi-carrier systems, in order to adapt to different channel environment condition, it will usually have between multiple protection Every the selection of ratio mode, in receivers, it must be determined that the length at the protection interval that current system selects The original position of determination OFDM symbol that could be correct.In multi-carrier systems, in order to correct connects Receiving signal, receiver must determine the original position of OFDM symbol, i.e. Symbol Timing.

DVB-T2 system is a ground system of digital television broadcast based on OFDM multi-transceiver technology, Its baseband signal is divided into superframe one by one, comprises most 255 T2 frames inside each superframe, Each T2 frame starts with a P1 sign bit, is followed by N number of P2 symbol, is multiple data symbols afterwards Number, wherein P2 symbol and data symbol have identical FFT length and protection interval (GI) length.

The structure of P1 symbol is fixed, and a length of 2048, it will be seen that part A is by signaling S1 in Fig. 2 After encoded with S2, being what 1K IFFT obtained, C and B is the first half of A and later half respectively The weighting of part replicates, and its product process is shown in Fig. 3.

The feature of P2 symbol be pilot tone than comparatively dense, pilot interval is multiple or the multiple of 6 of 3, by The length of FFT and the type decided of MISO.

The Symbol Timing of DVB-T2 system and mode detection method are typically [1] of Combined estimator, i.e. utilize The feature of the Cyclic Prefix of OFDM symbol, by sliding, relevant method obtains peak value, then judges two Interval between individual peak value, determines transmission mode and the length at protection interval with this, also completes simultaneously Symbol Timing.The principle of algorithm is as follows:

Definition auto-correlation function is:

$R_i\left(n\right)=\frac{1}{N_i}\underset{j=0}{\overset{\frac{N_i}{Q}-1}{\mathrm{\Σ}}}r(n-j)r^{*}(n-j-N_i)---\left(1\right)$

In formula (1), the length of N standard transmission model F FT, DVB-T has 2K and 8K two Select, and in DVB-T2, have 1K, 2K, 4K, 8K, 16K, 32K etc. 5 kinds selection.N/Q table Showing the length of GI, in DVB-T, Q has 4 kinds of values (4,8,16,32), and Q in DVB-T2 Then there are 7 kinds of values (4,8,16,32,128,128/19,256/19).

Corresponding auto-correlation function can be obtained as shown in Figure 6.

If Q arranges correctly, it is possible to judge transmission mode by calculating the distance between two peak values. From the above equation we can easily see, if GI has N kind to select in system, then in order to estimate The pattern of meter GI, it is necessary to do N-1 above-mentioned related operation.Due in DVB-T2 system, the shortest The length of GI be 1/128, in such a mode, relevant peak value can be the least, especially in channel disturbance In the case of very big, it is easy to erroneous judgement occurs.

In DVB-T2 system, Symbol Timing starts after frame synchronization.After frame synchronization locking, Can provide two parameters S1 and S2, S1 is the parameter of 3bit, and the system representing current is MISO pattern Or SISO pattern, and T2-BASE or T2-LITE pattern；S2 is the parameter of 4bit, comprises FFT SIZE and the information of part GI pattern.Owing to the timing position of frame synchronization exists error, so Needs carry out the most accurate timing synchronization.(S2 parameter is simply for the pattern of GI to be judged Tell the span of GI pattern).It should be noted that in DVB-T2 system, a total of seven Plant GI pattern, be 1/4,1/8,1/16,1/32,1/128,19/128,19/256 respectively, and DVB-T compares, and many Three models below, especially 1/128 pattern, length is the shortest, if using DVB-T Mode detection method, just have the highest detection error.

If by the Symbol Timing in existing DVB-T system and mode detection technology, needing at least two OFDM symbol carries out relevant judgement of sliding.And carrying out sliding relevant when, must be at least Do 6 such actions.From formula (1) it will be seen that under 32K and 16K pattern, because of Bigger for sliding window, operand in this way very big, this means that the COST of chip Can be bigger with POWER.On the other hand, due to because the shortest GI length is only 64 point (8K patterns Lower 1/128GI, or 1/32GI under 1K pattern), so peak value is the least after slip is relevant, many when occurring The when that footpath channel and frequency offset error being bigger, peak value will be inconspicuous, thus the mistake of big probability occurs Sentencing, namely the performance of chip can not be guaranteed.So the way of the Symbol Timing of DVB-T is the most Through being not applied in DVB-T2 system, need to make improvements.

Summary of the invention

The technical problem to be solved is, not enough for prior art, it is provided that a kind of DVB-T2 system The method of Symbol Timing correction in system, reduces the Symbol Timing in DVB-T system and mode detection method Operand and the time of capture, improve the accuracy of mode detection.

For solving above-mentioned technical problem, the technical solution adopted in the present invention is: in a kind of DVB-T2 system The method of Symbol Timing correction, the method is:

1) two parameters S1 provided after the locking of DVB-T2 frame synchronization and S2 is utilized to generate P1 symbol；

2) the I/Q two-way of above-mentioned P1 symbol is taken sign bit respectively, obtain sign synchronization correlated series Ssc_seq:

K=[0～1023],

Wherein $\mathrm{sign}\left[x\right]=\left\{\begin{array}{cc}+1,& x\&GreaterEqual;0\\ -1,& x<0\end{array}\right.,$ Represent the real part taking P1,Represent and take P1 Imaginary part；

3) sign synchronization correlated series ssc_seq does segmented plain be correlated with:

$\mathrm{corr}\_\mathrm{result}\left(l\right)=\underset{n=0}{\overset{\mathrm{seg}\_\mathrm{num}-1}{\mathrm{\&Sigma;}}}\left(\right|\underset{m=n*\mathrm{seg}\_\mathrm{len}}{\overset{(n+1)*\mathrm{seg}\_\mathrm{len}-1}{\mathrm{\&Sigma;}}}r(m+k)*\mathrm{ssc}\_\mathrm{seq}\left(m\right)\left|\right),$

Wherein, corr_result is segmented plain correlated results, after r (m) is frame synchronization locking Receiving signal, seg_num is the number of the segmentation that slip is relevant, and seg_len is slip phase Close the length of each section of segmentation, seg_len*seg_num=1024 ,-64≤l≤64；

4) according to the segmented plain associated expression in step 3), segmented plain correlated results is determined Maximum and position thereof, find first P2 symbol (comprising GI) r in reception signal_p2(s)；

5) utilization determines the pilot frequency design of P2 according to S1 and S2, and accordingly generates mode detection sequence Row md_seq0 (w)；

6) initial value GI=1/4 is set；

7) part that the tail length of mode detection sequence md_seq0 (w) is GI is moved head, Obtain:

$\mathrm{md}\_\mathrm{seq}\left(w\right)=\left\{\begin{array}{cc}\mathrm{md}\_\mathrm{seq}0(N-N*\mathrm{GI}+w)& 0\&le;w<\mathrm{GI}*N\\ \mathrm{md}\_\mathrm{seq}0(w-\mathrm{GI}*N)& \mathrm{GI}*N\&le;w<N\end{array}\right.$

Wherein, w=[0～N-1], N are the length of FFT；

8) to sequence md_seq and r_p2Doing segmented plain to be correlated with, sliding distance is 5:

$\mathrm{corr}\_\mathrm{result}\left(l\right)=\underset{n=0}{\overset{\mathrm{seg}\_\mathrm{num}-1}{\mathrm{\&Sigma;}}}\left(\right|\underset{m=n*\mathrm{seg}\_\mathrm{len}}{\overset{(n+1)*\mathrm{seg}\_\mathrm{len}-1}{\mathrm{\&Sigma;}}}{r}_{p2}(m+t)*\mathrm{md}\_\mathrm{seq}\left(m\right)\left|\right),$

Wherein, $\mathrm{seg}\_\mathrm{len}*\mathrm{seg}\_\mathrm{num}=\left\{\begin{array}{cc}N& N\&le;8192\\ 8192& N>8192\end{array}\right.,$ t=[0,1,2,3,4,5]

9) by the 5 of step 8) slip correlated results summations, the correlated results of current GI pattern is obtained；

10) respectively GI is set as other 6 kinds of patterns, i.e. GI=1/8,1/16,1/32,1/128,19/128,19/256, repeat 7) 9), obtain 7 phases Close result；

11) comparing the maximum of these 7 correlated results, the GI pattern of its correspondence is exactly mode detection Result；

12) in above-mentioned steps 11) mode detection results is corresponding 5 slide in relevant result, Maximizing, the position of described maximum is exactly the result of final correction symbol timing.

Compared with prior art, the had the beneficial effect that present invention of the present invention takes full advantage of The feature of the frame structure of DVB-T2 system, itself P1 and P2 symbol has point-device dependency, according to Symbol Timing position and the accuracy of the testing result of GI pattern that this obtains have obtained bright than method before Aobvious improvement；Related operation due to the present invention has only used the sign bit of sequence, thus is greatly simplified The accumulating operation that is multiplied, reduces cost；In sign synchronization, owing to the length of P1 signal only has 1024 Point, so, for all of FFT mode and GI pattern, this related operation has only to a length of 1024 The sequence of point.Because this sign synchronization starts after frame synchronization, thus slide relevant distance and The precision of frame synchronization is relevant, and according to the feature of P1 signal, this accuracy rating is [-64,64].Because During Xiang Guan, only take the sign bit of P1 symbol, thus avoid multiplying, only addition.Institute With, the operand that completing sign synchronization needs is 1024*128 complex addition operations, i.e. 262144 Additive operation；Same in mode detection process, only additive operation, the number of additive operation is: 8192*5*7*2=573440；The method of the present invention significantly reduces operand and computing cost, improves Mode detection accuracy.

Accompanying drawing explanation

Fig. 1 is DVB-T2 frame structure；

Fig. 2 is DVB-T2P1 symbolic construction；

Fig. 3 is that DVB-T2 generates P1 symbol flow chart；

Fig. 4 is the pilot frequency design (MISO pattern) of DVB-T2 P2 symbol；

Fig. 5 is the pilot frequency design (SISO pattern) of DVB-T2 P2 symbol；

Fig. 6 is the autocorrelation function graph that existing method obtains；

Fig. 7 is the PRBS generator structural representation generating scramble sequence；

Fig. 8 is sign synchronization flow chart of the present invention；

Fig. 9 is the PRBS Sequence generator schematic diagram generating PRBS Sequence；

Figure 10 is the inventive method flow chart.

Detailed description of the invention

In the digital television systems such as DVB-T2, synchronization module sign synchronization to be done is entered the most again Row mode detection and frequency deviation are estimated.The former is the accurate location finding signal frame and symbol, and the latter has been Further signal analysis and parsing, finds transmission mode and the frequency deviation of system.

[a] sign synchronization

1) S1 and S2 is utilized to generate P1 symbol.

After locking due to frame synchronization, can provide two parameters S1 and S2, the two parameter indicates currently The FFT mode of system employing and MISO pattern, which determine the frame head P1 symbol of DVB-T2 system Content and structure.

Table 1S1 and S2 modulating mode

From upper table it will be seen that S1 and S2 is the parameter of 3bit and 4bit respectively, correspond to 8 Hes respectively 16 possible sequences.Wherein corresponding for S1 sequence length is that the sequence length that 64bit, S2 are corresponding is 256bit.The generation process of P1 sequence is described below, refers to list of references [2].

First, select modulation sequence according to S1 and S2:

{mss_seq₀…mss_seq₃₈₃}={css_S1,css_S2,css_S1}

={css_S1,0…css_S1,63,css_S2,0…css_S2,255,css_S1,0…css_S1,63}

Wherein, mss_seq is the sequence of a length of 384bit, through follow-up conversion, modulation to 1K FFT Corresponding 384 sub-carrier positions on.It is produced by S1 and S2 sequence.S1 is the letter of 3bit Breath, has 8 kinds of selections, correspond to 64 bit information of every a line of S1 in table above respectively css_S1={css_S1,0…css_S1,63}.S2 is the information of 4bit, has 16 kinds of selections, correspond to respectively in table Information css of every a line 255bit of S2_S2={css_S2,0…css_S2,255}。

Second, modulation intelligence above is DBPSK and modulates:

mss_diff=DBPSK(mss_seq)

$\mathrm{mss}\_{\mathrm{diff}}_i=\left\{\begin{array}{cc}\mathrm{mss}\_{\mathrm{diff}}_{i-1}& ,\mathrm{mss}\_{\mathrm{seq}}_i=0\\ -\mathrm{mss}\_{\mathrm{diff}}_{i-1}& ,\mathrm{mss}\_{\mathrm{seq}}_i=1\end{array}\right.$

mss_diff_-1=+1

Wherein, mss_diff is the information of the mss_seq a length of 384bit after DBPSK modulates.

3rd, sequence scrambling above:

mss_scr=SCRAMBLING{mss_diff}

Scramble sequence PRBS generator as shown in Figure 7 produces, a length of 384bit, and scrambling PRBS is raw The generator polynomial grown up to be a useful person is: 1+x¹⁴+x¹⁵；Wherein x is 0 or 1, scrambled after sequence permissible Represent by equation below:

mss_scr_i=mss_diff_i×(1-2PRBS_i)

Wherein, PRBS_iIt is the bit (0 or 1) of current time PRBS generator output, correspond to DBPSK The current bit, i=[0～383] doing scrambling process in result.

4th, above-mentioned sequence is assigned on the subcarrier of 1K pattern according to fixing position.

5th, it is IFFT, obtains P1 symbol.

6th, the I/Q two-way of P1 symbol obtained above is taken sign bit respectively, obtains present invention needs Sign synchronization correlated series:

K=[0～1023]

Wherein $\mathrm{sign}\left[x\right]=\left\{\begin{array}{cc}+1,& x\&GreaterEqual;0\\ -1,& x<0\end{array}\right.,$ Represent the real part taking P1,Imaginary part for P1.

Sign synchronization flow chart is shown in Fig. 8.

2) segmented plain is correlated with

Owing to, carrying out sign synchronization when, carrier synchronization being had not been completed, system there is also certain The frequency deviation of degree, causes receiving between signal and sign synchronization correlated series and does not mates, if now direct Do slip relevant, peak value would not occur.Accordingly, it would be desirable to relevant way of sliding is improved, i.e. Segmented plain is correlated with.

$\mathrm{corr}\_\mathrm{result}\left(l\right)=\underset{n=0}{\overset{\mathrm{seg}\_\mathrm{num}-1}{\mathrm{\&Sigma;}}}\left(\right|\underset{m=n*\mathrm{seg}\_\mathrm{len}}{\overset{(n+1)*\mathrm{seg}\_\mathrm{len}-1}{\mathrm{\&Sigma;}}}r(m+k)*\mathrm{ssc}\_\mathrm{seq}\left(m\right)\left|\right)$

Wherein, r (m) is the reception signal after frame synchronization, and seg_num is the number of the segmentation that slip is relevant, Seg_len is the length of each section.Seg_len*seg_num=1024 ,-64≤l≤64.

3) detection peak value

Owing to the error of frame synchronization is up to 1024 points, so above-mentioned slides relevant length preferably at most 1024.Do slip relevant while, need to detect maximum, when slide relevant end when, also Just having obtained maximum and the position thereof of correlated results, this position is exactly the result of sign synchronization.

[b] mode detection and sign synchronization correction

Due in DVB-T2 system, the beginning of each frame has a P2 symbol, and the feature of P2 symbol Be pilot tone than comparatively dense, pilot frequency design see Figure4. we can utilize P2 symbol time-domain information come Carry out mode detection and sign synchronization correction, owing to the dependency of P2 signal is relatively good, so, at 16K With under 32K pattern, we only take the length of 8K, say, that during mode detection, The longest sequence is 8K length.Specifically comprise the following steps that

1) utilize P2 symbol pilot generation mode detection sequence:

A) the PRBS Sequence generator be given by Fig. 9, generates PRBS Sequence p2_prbs0 (p), Wherein p=[0～kmax-1], kmax is that under current transmission mode, in P2 symbol, effectively son carries The number of ripple,；The generator polynomial of the PRBS generator shown in Fig. 9 is: x¹¹+x²+ 1, Wherein x is 0 or 1；

B) according to current transmission mode, i.e. FFT or MISO pattern determines between P2 symbol pilot tone Every Dx；

C) by the position zero setting of the integral multiple of non-Dx in above-mentioned p2_prbs1 (p) sequence；

P=[0～kmax-1]

D) sequence p2_prbs1 (p) is circulated displacement, obtains sequence p2_prbs2 (w)

$p2\_\mathrm{prbs}2\left(w\right)=\left\{\begin{array}{cc}p2\_\mathrm{prbs}1(w+\frac{k\max -1}{2}),& w\&le;\frac{k\max -1}{2}\\ 0,& \frac{k\max -1}{2}<w\&le;N-\frac{k\max -1}{2}\\ p2\_\mathrm{prbs}1(w+\frac{k\max -1}{2}-N),& N-\frac{k\max -1}{2}<w\&le;N-1\end{array}\right.$

E) sequence after step d) processes is done IFFT conversion, obtain the time domain of P2 symbol pilot tone Sequence p2_pilot3 (w), w=[0～N-1]；

F) p2_pilot3 (w) is taken sign bit, obtains mode detection sequence:

Wherein, w=[0～N-1], N are the length of FFT.

2) mode detection

Owing to the error of sign synchronization is+-2, so, the relevant window that slides is 5；

A) GI=1/4 is set；

B) part that tail length is GI of mode detection sequence md_seq0 is moved head, Arrive:

$\mathrm{md}\_\mathrm{seq}\left(w\right)=\left\{\begin{array}{cc}\mathrm{md}\_\mathrm{seq}0(N-N*\mathrm{GI}+w)& 0\&le;w<\mathrm{GI}*N\\ \mathrm{md}\_\mathrm{seq}0(w-\mathrm{GI}*N)& \mathrm{GI}*N\&le;w<N\end{array}\right.$

Wherein, w=[0～N-1], N are the length of FFT；

C) do segmented plain to be correlated with:

$\mathrm{corr}\_\mathrm{result}\left(l\right)=\underset{n=0}{\overset{\mathrm{seg}\_\mathrm{num}-1}{\mathrm{\&Sigma;}}}\left(\right|\underset{m=n*\mathrm{seg}\_\mathrm{len}}{\overset{(n+1)*\mathrm{seg}\_\mathrm{len}-1}{\mathrm{\&Sigma;}}}{r}_{p2}(m+t)*\mathrm{md}\_\mathrm{seq}\left(m\right)\left|\right),$

Wherein, $\mathrm{seg}\_\mathrm{len}*\mathrm{seg}\_\mathrm{num}=\left\{\begin{array}{cc}N& N\&le;8192\\ 8192& N>8192\end{array}\right.,$ t=[0,1,2,3,4,5]

D) by 5 correlated results summations, the correlated results of current GI pattern is obtained.Retain former 5 simultaneously Point correlated results；

E) respectively GI is set as other 6 kinds of patterns, repeats b), c), d), obtain 7 relevant knots Really；

F) comparing the maximum of these 7 correlated results, the GI pattern of its correspondence is exactly mode detection Result.

3) Symbol Timing correction

Above-mentioned mode detection results corresponding 5 slide in relevant result, maximizing, Its position, it is simply that we need the result of the final correction symbol timing obtained.

The inventive method flow chart is shown in Figure 10.

Sign synchronization in the present invention starts on the basis of frame synchronization.The frame synchronization of DVB-T2 system Although signaling S1 and S2 more accurately can be given, but there is the biggest error its location, especially At channel condition than under relatively rugged environment, the position location of usual frame synchronization and real FFT energy are Big the window's position has bigger deviation.Determine accordingly, it would be desirable to sign synchronization further corrects Position.

The present invention makes full use of the feature of P1 symbol in DVB-T2 system, assists and carries out sign synchronization. Owing to the signal-noise ratio threshold of the coding and decoding of P1 symbol is the lowest, therefore, as long as frame synchronization can be of poor quality After orienting the position of P1 symbol, what system just can be correct solves signaling S1 and S2.Present invention profit Use the two parameter, use the generation method (see figure 3) of transmitter P1 symbol, structure again at receiving terminal Produce P1 symbol.Then the signal after this signal and the frame synchronization received is done slip relevant, by There is frequency deviation in the signal received, relevant needs of sliding the most herein does the relevant, in phase of segmentation Close the position finding peak value in result, thus find the original position of P2 symbol.

Do above-mentioned slip relevant during, in order to reduce area, the P1 symbol regenerated can only take Sign bit, this simplification has no effect on the dependency of P1 symbol.

The present invention pilot tone (see figure 4) of P2 symbol, is calculated by IFFT, generates local sequence, Because the pilot frequency mode of P2 it is known that so, this sequence can be in being stored in advance in ROM.In order to Reducing hardware spending, we the most only take its sign bit.Owing to the dependency of P2 signal is relatively good, so, Under 16K and 32K pattern, we only take the length of 8K, say, that in mode detection During, the longest sequence is 8K length.

Do slip with this P2 symbol with reception data relevant, owing to there is also frequency deviation in system, therefore still need to The slip of segmentation to be done is correlated with.Because in the step of sign synchronization, timing error+-2, therefore I Need to do the slip of 5 and be correlated with, for doing 5 again after each GI pattern transposition P2 sequence respectively The slip of point is correlated with.The result the most altogether obtained is the array of 5 × 7.

First, 5 slip correlated results of each GI pattern are done weighted average by us, then find 7 Maximum inside numerical value, i.e. we need the GI estimated to the GI of its correspondence.Then, at corresponding GI The relevant result of 5 slips in, choose the maximum, this point is exactly the timing position of accurate sign synchronization Put.

By above-mentioned method, we are while obtaining GI estimation accurately, also accurate locking The original position of OFDM symbol.

The list of references that the present invention uses:

[1]Optimum receiver design for OFDM-based broadband transmission-part II, Michael Speth,IEEE2001；

[2]ETSI EN 302 755 V1.2.1,Digital Video Broadcasting(DVB);Frame structure channel coding and modulation for a second generation digital terrestrial television broadcasting system(DVB-T2)。

Claims (2)

1. the method for Symbol Timing correction in a DVB-T2 system, it is characterised in that the method is:

1) utilize two parameters S1 provided after the locking of DVB-T2 frame synchronization and S2 to generate to receive Machine P1 symbol under current transmission mode；

2) the I/Q two-way of above-mentioned P1 symbol is taken sign bit respectively, obtain sign synchronization correlated series Ssc_seq:

Wherein $sign\&lsqb;x\&rsqb;=\left\{\begin{array}{cc}+1,& x\&GreaterEqual;0\\ -1,& x<0\end{array}\right.,$ Represent the real part taking P1,Expression takes The imaginary part of P1；

3) sign synchronization correlated series ssc_seq does segmented plain be correlated with:

$corr\_result\left(h\right)=\underset{n=0}{\overset{seg\_num-1}{\&Sigma;}}\left(\right|\underset{m=n*seg\_len}{\overset{(n+1)*seg\_len-1}{\&Sigma;}}r(m+h)*ssc\_seq\left(m\right)\left|\right),$

Wherein, corr_result is segmented plain correlated results, after r (m) is frame synchronization locking Reception signal, seg_num is the number of relevant segmentation of sliding, and seg_len is for sliding The length of dynamic each section of relevant segments, seg_len*seg_num=1024, -64≤h≤64；H is sliding distance；

4) according to step 3) in segmented plain associated expression, determine segmented plain correlated results Maximum and position, find and receive first P2 symbol comprising GI part in signal Number r_p2(s)；GI is the protection interval of P2 symbol；

5) utilization determines the pilot frequency design of P2 according to S1 and S2, and accordingly generates mode detection sequence Row md_seq0 (w)；

6) initial value GI=1/4 is set；

7) part that the tail length of mode detection sequence md_seq0 (w) is GI is moved head, Obtain:

$md\_seq\left(w\right)=\left\{\begin{array}{cc}md\_seq0(N-N*GI+w)& 0\&le;w<GI*N\\ md\_seq0(w-GI*N)& GI*N\&le;w<N\end{array}\right.$

Wherein, w=[0～N-1], N are the length of FFT；

8) to sequence md_seq and r_p2Doing segmented plain to be correlated with, sliding distance is 5:

$corr\_result\left(h\right)=\underset{n=0}{\overset{seg\_num-1}{\&Sigma;}}\left(\right|\underset{m=n*seg\_len}{\overset{(n+1)*seg\_len-1}{\&Sigma;}}{r}_{p2}(m+t)*md\_seq\left(m\right)\left|\right),$

Wherein, $seg\_len*seg\_num=\left\{\begin{array}{cc}N& N\&le;8192\\ 8192& N>8192\end{array}\right.,t=\&lsqb;0,1,2,3,4,5\&rsqb;$

9) by step 8) the summation of 5 slip correlated results, obtain the relevant knot of current GI pattern Really；

10) respectively GI is set as other 6 kinds of patterns, i.e. GI=1/8,1/16,1/32,1/128,19/128,19/256, repeat 7) 9), obtain 7 Correlated results；

11) comparing the maximum of these 7 correlated results, the GI pattern of its correspondence is exactly mode detection Result；

12) in above-mentioned steps 11) mode detection results is corresponding 5 slide in relevant result, Maximizing, the position of described maximum is exactly the result of final Symbol Timing correction.

The method of Symbol Timing correction in DVB-T2 system the most according to claim 1, its feature exists In, described step 5) in, utilize P2 symbol pilot generation mode detection sequence md_seq0 (w) Process be:

1) by the PRBS Sequence generator of DVB-T2, PRBS Sequence is generated P2_prbs0 (p), wherein p=[0～k max-1], k max is current transmission mode The number of effective subcarrier in lower P2 symbol；The generation of described PRBS Sequence generator is many Xiang Shiwei: x¹¹+x²+1；Wherein x is 0 or 1；

2) P2 symbol pilot tone is determined according to current transmission mode MISO pattern and FFT length Interval D x；

3) by the position zero setting of the integral multiple of non-Dx in above-mentioned p2_prbs0 (p) sequence；

P=[0～k max-1]

4) sequence p2_prbs1 (p) is circulated displacement, obtains sequence p2_prbs2 (w)

$p2\_prbs2\left(w\right)=\left\{\begin{array}{cc}p2\_prbs1(w+\frac{k\max -1}{2}),& w\&le;\frac{k\max -1}{2}\\ 0,& \frac{k\max -1}{2}<w\&le;N-\frac{k\max -1}{2}\\ p2\_prbs1(w+\frac{k\max -1}{2}-N),& N-\frac{k\max -1}{2}<w\&le;N-1\end{array}\right.$

5) to through step 4) process after sequence be IFFT conversion, obtain P2 symbol pilot tone time Territory sequence p2_pilot3 (w), w=[0～N-1]；

6) p2_pilot3 (w) is taken sign bit, obtains mode detection sequence:

Wherein, w=[0～N-1], N are the length of FFT.