Frame header decoding method of digital satellite TV signal in ultra-low signal-to-noise ratio environment

Abstract

The present invention discloses a frame header decoding method for digital satellite TV signals in an ultra-low signal-to-noise ratio environment. In the method of the present invention, if there is spectrum inversion in the signal at the transmitting end, the imaginary part value of the PLS signaling information to be decoded is firstly inverted; the original frame header signaling coding value B=[b 0 ,b 1 ,b 2 ,b 3 ,b 4 ,b 5 ,b 6 ,b 7 ] 0 is 1 or 0, the output of the differential signal sequence is XORed to calculate the cross-select sequence, and the five-level butterfly calculation is performed after decoding preprocessing to obtain [b 2 ,b 3 ,b 4 ,b 5 ,b 6 ,b 7 ]; Finally, after hypothesis testing, we get b 1 is 1 or 0. The method of the present invention has a simple structure and can be directly implemented in a chip without modifying the encoding method of the transmitting end. It can stably decode in a lower signal-to-noise ratio environment and support the DVB-S2X mode.

Description

Technical Field

[0001] The invention belongs to the technical field of digital satellite television broadcasting, and in particular relates to a method for completing satellite signal frame header decoding in an ultra-low signal-to-noise ratio environment as low as -6dB to -10dB. Background Art

[0002] In the field of digital satellite TV broadcasting technology, in order to make the broadcast signal have a wider coverage, a feasible technical solution is to design the LDPC standard decoding threshold of the transmitted signal to be very low. For example, the European satellite digital video broadcasting standard DVB-S2X has designed a 4 / 15 code rate BPSK signal, and its theoretical signal-to-noise ratio threshold for normal operation is -4.9dB, that is, the noise power is about 3 times the signal power, and the target signal is completely submerged in the noise. Such a low signal-to-noise ratio makes it difficult for the traditional frame header signaling encoding method based on RM code as the encoding standard to continuously and correctly decode the frame header signaling value. The frame header signaling value is important information for LDPC encoding to correctly decode the current frame. It contains information such as the length, encoding method, code rate, and pilot frequency of the current frame. An error in frame header decoding means that it cannot be correctly decoded, which will also lead to the inability to demodulate the TV program normally. After the frame header is decoded correctly, continuous frame synchronization operations need to be completed to ensure that the subsequent LDPC frame decoding can be carried out continuously. LDPC continuous decoding can ensure uninterrupted video playback. Therefore, frame header decoding and synchronization is an important technology in the field of digital satellite TV broadcasting technology.

[0003] Generally, there are several technical solutions for identifying frame headers: Patent CN108011651B adds a part of unique word information to the original frame header information, increases redundancy, and makes frame header synchronization easier. However, for the established DVB-S2X standard, it is impossible to add additional redundant information to the original signal for frame header synchronization or decoding. Patent CN101414848B merges the acquired frame control header coding information and then decodes and parses the merged frame control header coding information, which is equivalent to merging multiple frames of information. However, in the case of DVB-S2X containing VCM, the frame header information of each frame may change. Obviously, this method is not applicable. The signal-to-noise ratio threshold designed by patent CN110034914B is -2.35dB, which is difficult to work normally in an ultra-low signal-to-noise ratio environment as low as -6dB, and this method is only applicable to DVB-S2 signals. Patent CN107659354A uses a longer frame header sequence to enable it to withstand harsh rainy environments, but in the field of digital satellite television broadcasting technology, the originating frame header sequence cannot be modified. Summary of the invention

[0004] The purpose of the present invention is to provide a frame header decoding method for digital satellite TV signals in an ultra-low signal-to-noise ratio environment as low as -6dB, so as to solve the problem of identifying the original coding value of the frame header signaling in the ultra-low signal-to-noise ratio environment.

[0005] The present invention specifically is:

[0006] Step (1) The receiver receives the digital broadcasting television signal forwarded by the satellite through the antenna, and obtains the complete 90 symbol information R of the frame header through frequency conversion, analog-to-digital conversion, and demodulation operations. plh =[R sof ,R pls ]=[r1,r2,...,r 90 ]; the symbol information of the first 26 moments represents the frame start signal R sof =[r1,r2,...,r 26 ], is the known sequence at the transmitting end, and the symbol information at the last 64 moments is the PLS signaling information to be decoded R pls =[r 27 ,r 28 ,...,r 90 ], the symbol information at each moment is complex;

[0007] Step (2) spectrum flip correction;

[0008] Calculate the previous sequence and value T sof =∑(Re(R sof )Im(SR sof )-Im(R sof )Re(SR sof )); where S matrix is ​​the first-order Toeplitz matrix of the unit matrix, Re(·) represents the real part of the sequence, and Im(·) represents the imaginary part of the sequence; if T sof ≥0, it is determined that the signal sent by the transmitter does not have spectrum inversion, otherwise it is determined that there is spectrum inversion; if there is spectrum inversion, it is necessary to adjust R pls The imaginary part of is negated;

[0009] Step (3) Demodulation by differential calculation;

[0010] According to the frame header coding rules of DVB-S2X, R pls =HI IQ64 I scr I b (1-2G′B)+N; where H is the channel state, N is the random Gaussian noise of the channel environment; I IQ64 is the bit-to-IQ mapping matrix, which is determined by the bit->IQ encoding rule described in the DVB-S2X standard; I bis the known cross selection matrix, I scr is the known frame header scrambling matrix, G′ is the known RM code encoding matrix, B is the original frame header signaling encoding value B = [b0, b1, b2, b3, b4, b5, b6, b7];

[0011] Take R pls =[r 27 ,r 28 ,...,r 90 ] two consecutive values, perform differential multiplication and subtraction operations, and solve the differential signal sequence d IQ64 =Re(r t )Im(r t+1 )-Im(r t )Re(r t+1 ), 26≤t≤89; for 64 moments d IQ64 Perform a sum operation. If the sum is greater than 0, b0 in B is judged to be 1, otherwise b0 is judged to be 0;

[0012] Step (4) for d IQ64 According to the known sequence I scr XOR them in chronological order, and the output after XOR is d scr ;

[0013] Step (5) extract the d of two adjacent moments in chronological order scr Data, denoted as d scr (t) and d scr (t+1), 1≤t≤63, according to the known cross selection matrix I b Compute solution crossover selection sequence Decision function

[0014] Step (6) decoding preprocessing;

[0015] All the moments d b The value is stored in the register, and the solution cross-select sequence d is selected according to the first row parameter of the G′ matrix b Perform descrambling operation, that is, directly b According to the first row parameters of the G′ matrix, XOR them in chronological order, and the output sequence after XOR is d G1 ; for d G1 The data of the even-numbered moments are inverted and superimposed on the odd-numbered moments to obtain the data output d f =[d f (1),d f (2),...,d f (32)],d f (i) is the preprocessed data output at the i-th moment, i=1, 2, ..., 32;

[0016] Step (7) preprocesses the data output d at each moment f (i) Perform five-level butterfly calculation: Butterfly calculation output Where W is the Hadamard matrix;

[0017] Step (8) If 32 butterfly calculations output d w If the maximum value in (k) is greater than the set threshold, the maximum value d w The number k′ of (k′) is argmax{d w (k)} is encoded as [b2,b3,b4,b5,b6] in B = [b0,b1,b2,b3,b4,b5,b6,b7], and b7 = 1;

[0018] If the 32 butterfly calculations output d w If the maximum value in (k) is less than or equal to the set threshold, then mark b7 = 0, remove the descrambling sequence preprocessed in step (6), do not perform descrambling, repeat step (7), and calculate the 32 butterfly outputs d w The binary code corresponding to the sequence number of the maximum value in (k) is taken as [b2,b3,b4,b5,b6] in B=[b0,b1,b2,b3,b4,b5,b6,b7];

[0019] Step (9) synthesizes the sequences [b0,1,b2,b3,b4,b5,b6,b7] and [b0,0,b2,b3,b4,b5,b6,b7] into 64-bit data S according to the coding rules of DVB-S2X. pls (1) and S pls (0), S pls (1) and S pls (0) respectively with the symbol information R of the last 64 moments received pls Do the correlation calculation and get the correlation peak value p1=∑S pls (1) R pls and p0=∑S pls (0)·R pls If p1>p0, then b1=1, otherwise b1=0. So far, all [b0,b1,b2,b3,b4,b5,b6,b7] have been parsed.

[0020] The method of the present invention first constructs a frame header coding expression, and designs a decoding method based on the expression. The decoding method of the present invention has a simple structure and can be directly implemented in a chip without a complex algorithm. On the other hand, the present invention does not need to modify the encoding method of the transmitter, and can stably decode in a lower signal-to-noise ratio environment under the existing 90-bit symbol header length, and can support the DVB-S2X mode. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is the encoding method of DVB-S2X frame header signaling;

[0022] Figure 2 It is a system block diagram of a digital broadcast television receiver;

[0023] Figure 3 is a system block diagram of the method of the present invention;

[0024] Figure 4 It is a schematic diagram of the differential operation in the method of the present invention;

[0025] Figure 5 Schematic diagram of descrambling operation in the method of the present invention;

[0026] Figure 6 This is a decoding structure diagram in the method of the present invention. DETAILED DESCRIPTION

[0027] For the frame header signal of DVB-S2X, its coding structure is as follows: Figure 1 As shown, B = [b0, b1, b2, b3, b4, b5, b6, b7] represents the 8-bit frame header signaling information (PLS) of the transmitter. The 7-bit information of PLS ​​bit [0:6] is encoded by the (32,7) RM code matrix to generate 32-bit data [y1, y2, y3, ..., y 32 ],y i ∈{0,1},1≤i≤32. Then y i XOR operation is performed with bit [7] of PLS ​​and inserted into the (32,7) RM code data to obtain the (64,8) code. The (64,8) coded data is then modulated by differential BPSK and sent to the satellite via the antenna, and the information is broadcast to the client via the satellite transponder. Since the satellite transponder channel and the air interface channel will cause a lot of noise to be superimposed on the PLS signaling information, in the worst case, the noise power may be stronger than the signal power. This method describes how to recover the PLS information in an ultra-low signal-to-noise ratio environment. The frame header decoding method of the digital satellite TV signal in an ultra-low signal-to-noise ratio environment is as follows: Figure 2 and 3 As shown:

[0028] Step (1) The receiver receives the digital broadcasting and television signal forwarded by the satellite through the antenna, sends it to the tuner to complete the frequency conversion operation, and then sends it to the analog-to-digital converter to complete the conversion of the analog signal to the digital signal, and completes the demodulation operation in the digital signal domain, including modules such as automatic gain control; then sends it to the timing recovery module, and after the recovery is completed, it is sent to the frame header synchronizer to extract the frame synchronization information, and completely obtain the 90 symbol information R of the frame header plh =[R sof ,R pls ]=[r1,r2,...,r 90 ]; the symbol information of the first 26 moments represents the frame start signal R sof =[r1,r2,...,r 26 ], is the known sequence at the transmitting end, and the symbol information at the last 64 moments is the PLS signaling information to be decoded R pls =[r 27 ,r 28 ,...,r 90 ], the symbol information at each moment is complex.

[0029] Step (2) spectrum flip correction;

[0030] Calculation sequence T = Re (R plh )Im(SR plh )-Im(R plh )Re(SR plh ), if the signal sent by the transmitter has spectrum inversion, then the inversion sequence T can be obtained by substituting the imaginary part into the above formula inv = -T, that is, the peak direction of the differential sequence with spectrum inversion is opposite to that of the differential sequence without spectrum inversion. sof =[r1,r2,...,r 26 ] is the known sequence at the start, that is, calculate the previous sequence and value T sof =∑(Re(R sof )Im(SR sof )-Im(R sof )Re(SR sof )); where S matrix is ​​the first-order Toeplitz matrix of the unit matrix, Re(·) represents the real part of the sequence, and Im(·) represents the imaginary part of the sequence. If T sof ≥0, it is determined that the signal sent by the transmitter does not have spectrum inversion, otherwise it is determined that there is spectrum inversion. If there is spectrum inversion, it is necessary to adjust R pls The imaginary part value of is inverted to achieve the purpose of spectrum flip correction, otherwise the spectrum flip correction operation is not required.

[0031] Step (3) Figure 4 As shown, the differential calculation is used to modulate;

[0032] According to the frame header coding rules of DVB-S2X, R pls =HI IQ64 I scr I b (1-2G′B)+N; where H is the channel state, which is determined by the real satellite channel, and N is the random Gaussian noise of the channel environment; I IQ64 is the bit-to-IQ mapping matrix, which is determined by the bit->IQ encoding rule described in the DVB-S2X standard; I b is the known cross selection matrix, I scr is the known frame header scrambling matrix, I scr =[011100011001110110000011110010010101001101000010001011011111101], G′ is the known RM code coding matrix, and B is the original frame header signaling coding value B = [b0, b1, b2, b3, b4, b5, b6, b7].

[0033] By R pls From the coding expression, we can know that to decode the original frame header signaling code value B, we only need to perform the inverse operation. That is, the PLS signaling information R to be decoded pls Multiply by the mapping matrix I IQ64 The inverse matrix of I IQ64 The characteristics of the inverse matrix, perform the following steps: Take R pls =[r 27 ,r 28 ,...,r 90 ] two consecutive values, perform differential multiplication and subtraction operations, and solve the differential signal sequence d IQ64 =Re(r t )Im(r t+1 )-Im(r t )Re(r t+1 ), 26≤t≤89, Figure 4 Middle -1 Indicates a delay of one moment; for 64 moments d IQ64 Perform a sum operation, if the sum is greater than 0, b0 in B is judged to be 1, otherwise b0 is judged to be 0.

[0034] Step (4) Figure 5 As shown, for 64 moments d IQ64 Perform descrambling operation, that is, directly IQ64 According to the known sequence I scr XOR them in chronological order, and the output after XOR is d scr .

[0035] Step (5) extract the d of two adjacent moments in chronological order scr Data, denoted as d scr (t) and d scr (t+1), 1≤t≤63, according to the known cross selection matrix I b Compute solution crossover selection sequence Decision function

[0036] Step (6) Figure 6 As shown, decoding preprocessing;

[0037] All the moments d b The value is stored in the register, and the solution cross-select sequence d is selected according to the first row parameter of the G′ matrix b Perform descrambling operation, that is, directly b According to the first row parameters of the G′ matrix, XOR them in chronological order, and the output sequence after XOR is d G1 ; for d G1 The data of the even-numbered moments are inverted and added to the odd-numbered moments, that is, the d of the 64 moments b The data is calculated as 32 data d f =[d f (1),d f (2),...,d f (32)],d f (i) is the preprocessed data output at the i-th moment, i=1,2,…,32.

[0038] Step (7) Figure 6 As shown, the preprocessed data output d at each moment f (i) Perform five-level butterfly calculation;

[0039] Butterfly calculation output Where W is the Hadamard matrix.

[0040] The nth output data d of the first-stage butterfly computing unit w (1,n)=d f (n)+d f (n+16), n=1,2,…,16, the mth output data d w (1,m)=d f (m-16)-d f (m), m = 17, 18, ..., 32;

[0041] The n1th output data d of the second-stage butterfly computing unit w (2,n1)=d w (1,n1)+d w(1,n1+8), n1=1,2,…,8, the n2th output data d w (2,n2)=d w (1,n2)-d w (1,n2+8), n2=9,10,…,16, the n3th output data d w (2,n3)=d w (1,n3)+d w (1,n3-16), n3=17,18,…,24, the n4th output data d w (2,n4)=d w (1,n4)-d w (1,n4-16), n4=25,26,…,32. And so on, overlapping calculations are performed in sequence, and the fifth-level butterfly calculation output is used as the output d of this step w (k).

[0042] Step (8) If 32 butterfly calculations output d w If the maximum value in (k) is greater than the set threshold, the maximum value d w The number k′ of (k′) is argmax{d w (k)} is encoded as [b2, b3, b4, b5, b6] in B = [b0, b1, b2, b3, b4, b5, b6, b7], and b7 = 1.

[0043] If the 32 butterfly calculations output d w If the maximum value in (k) is less than or equal to the set threshold, then mark b7 = 0, remove the descrambling sequence preprocessed in step (6), do not perform descrambling, repeat step (7), and calculate the 32 butterfly outputs d w The binary code corresponding to the serial number of the maximum value in (k) is taken as [b2,b3,b4,b5,b6] in B=[b0,b1,b2,b3,b4,b5,b6,b7].

[0044] Step (9) has solved [b0, b2, b3, b4, b5, b6, b7] so far, and only b1 has not been solved correctly.

[0045] Step (9) synthesizes the sequences [b0,1,b2,b3,b4,b5,b6,b7] and [b0,0,b2,b3,b4,b5,b6,b7] into 64-bit data S according to the coding rules of DVB-S2X. pls (1) and S pls (0), S pls (1) and S pls (0) respectively with the symbol information R of the last 64 moments receivedpls Do the correlation calculation and get the correlation peak value p1=∑S pls (1) R pls and p0=∑S pls (0)·R pls If p1>p0, then b1=1, otherwise b1=0. So far, all [b0,b1,b2,b3,b4,b5,b6,b7] have been parsed.

Claims

1. A frame header decoding method for digital satellite TV signals in an ultra-low signal-to-noise ratio environment, characterized in that: Step (1) The receiver receives the digital broadcasting television signal forwarded by the satellite through the antenna, and obtains the complete 90 symbol information R of the frame header through frequency conversion, analog-to-digital conversion, and demodulation operations. plh =[R sof ,R pls ]=[r1,r2,...,r 90 ]; the symbol information of the first 26 moments represents the frame start signal R sof =[r1,r2,...,r 26 ], is the known sequence at the transmitting end, and the symbol information at the last 64 moments is the PLS signaling information to be decoded R pls =[r 27 ,r 28 ,...,r 90 ], the symbol information at each moment is complex; Step (2) spectrum flip correction; Calculate the previous sequence and value T sof =∑(Re(R sof )Im(SR sof )-Im(R sof )Re(SR sof )); where S matrix is ​​the first-order Toeplitz matrix of the unit matrix, Re(·) represents the real part of the sequence, and Im(·) represents the imaginary part of the sequence; if T sof ≥0, it is determined that the signal sent by the transmitter does not have spectrum inversion, otherwise it is determined that there is spectrum inversion; if there is spectrum inversion, it is necessary to adjust R pls The imaginary part of is negated; Step (3) Demodulation by differential calculation; According to the frame header coding rules of DVB-S2X, R pls =HI IQ64 I scr I b (1-2G′B)+N; where H is the channel state, N is the random Gaussian noise of the channel environment; I IQ64 is the bit-to-IQ mapping matrix, which is determined by the bit->IQ encoding rule described in the DVB-S2X standard; I b is the known cross selection matrix, I scr is the known frame header scrambling matrix, G′ is the known RM code encoding matrix, B is the original frame header signaling encoding value B = [b0, b1, b2, b3, b4, b5, b6, b7]; Take R pls =[r 27 ,r 28 ,...,r 90 ] two consecutive values, perform differential multiplication and subtraction operations, and solve the differential signal sequence d IQ64 =Re(r t )Im(r t+1 )-Im(r t )Re(r t+1 ), 26≤t≤89; for 64 moments d IQ64 Perform a sum operation. If the sum is greater than 0, b0 in B is judged to be 1, otherwise b0 is judged to be 0; Step (4) for d IQ64 According to the known sequence I scr XOR them in chronological order, and the output after XOR is d scr ; Step (5) extract the d of two adjacent moments in chronological order scr Data, denoted as d scr (t) and d scr (t+1), 1≤t≤63, according to the known cross selection matrix I b Compute solution crossover selection sequence Decision function Step (6) decoding preprocessing; All the moments d b The value is stored in the register, and the solution cross-select sequence d is selected according to the first row parameter of the G′ matrix b Perform descrambling operation, that is, directly b According to the first row parameters of the G′ matrix, XOR them in chronological order, and the output sequence after XOR is d G1 ; for d G1 The data of the even-numbered moments are inverted and superimposed on the odd-numbered moments to obtain the data output d f =[d f (1),d f (2),...,d f (32)],d f (i) is the preprocessed data output at the i-th moment, i=1, 2, ..., 32; Step (7) preprocesses the data output d at each moment f (i) Perform five-level butterfly calculation: Butterfly calculation output Where W is the Hadamard matrix; Step (8) If 32 butterfly calculations output d w If the maximum value in (k) is greater than the set threshold, the maximum value d w The number k′ of (k′) is argmax{d w (k)} is encoded as [b2,b3,b4,b5,b6] in B = [b0,b1,b2,b3,b4,b5,b6,b7], and b7 = 1; If the 32 butterfly calculations output d w If the maximum value in (k) is less than or equal to the set threshold, then mark b7 = 0, remove the descrambling sequence preprocessed in step (6), do not perform descrambling, repeat step (7), and calculate the 32 butterfly outputs d w The binary code corresponding to the sequence number of the maximum value in (k) is taken as [b2,b3,b4,b5,b6] in B=[b0,b1,b2,b3,b4,b5,b6,b7]; Step (9) synthesizes the sequences [b0,1,b2,b3,b4,b5,b6,b7] and [b0,0,b2,b3,b4,b5,b6,b7] into 64-bit data S according to the coding rules of DVB-S2X. pls (1) and S pls (0), S pls (1) and S pls (0) respectively with the symbol information R of the last 64 moments received pls Do the correlation calculation and get the correlation peak value p1=∑S pls (1) R pls and p0=∑S pls (0)·R pls If p1>p0, then b1=1, otherwise b1=0. So far, all [b0,b1,b2,b3,b4,b5,b6,b7] have been parsed.

Images