CN102386932A - LDPC code constitution method - Google Patents
LDPC code constitution method Download PDFInfo
- Publication number
- CN102386932A CN102386932A CN2011103414024A CN201110341402A CN102386932A CN 102386932 A CN102386932 A CN 102386932A CN 2011103414024 A CN2011103414024 A CN 2011103414024A CN 201110341402 A CN201110341402 A CN 201110341402A CN 102386932 A CN102386932 A CN 102386932A
- Authority
- CN
- China
- Prior art keywords
- matrix
- row
- col
- sign indicating
- indicating number
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Abstract
The invention discloses a constitution method for two LDPC codes with different check matrix structures. The method comprises the steps as follows: firstly, the structure of the rarefaction check matrix H1 of the first LDPC code is confirmed; secondly, the parallel degree P and parameter n of an LDPC code encoder are selected according to the P and n search parameter combination (wrow, wco1 and k2); thirdly, according to the searched combination (wrow, wco1 and k2), a mother board matrix C is constructed; fourthly; a matrix Hi and j replaces the corresponding elements of the motherboard matrix C so as to obtain the check matrix H1of the first LDPC code; and fifthly, and the second check matrix H2 is obtained through ordering the rows and the lines of the H1. According to the invention, the first LDPC code with linear coding complexity and low required memory capacity can be constructed through simple search; the second LDPC code can be obtained through ordering the lines and the rows of the rarefaction check matrix of the first LDPC code; the first LDPC code with the same coed rate but different code lengths can be gained through simple replacing; and the design difficulty is greatly reduced.
Description
Technical field
The present invention relates to the communications field, particularly the building method of LDPC sign indicating number.
Background technology
Communication system adopts channel coding technology to guarantee the reliability in noisy communication channel communication usually.Such as, in satellite communication system, because there are a large amount of noise sources in geography and Effect of Environmental.These communication channels have its theoretic maximum communication capacity (just famous shannon limit), and this capacity can use the bit rate (bps) under the specific signal to noise ratio (snr) condition to represent.Therefore, the design of forward error correction (FEC) is exactly in order to pursue the maximum bit rate near shannon limit.Wherein a kind of coding near shannon limit is exactly low-density checksum (LDPC) sign indicating number.
The LDPC sign indicating number is a kind of linear block codes; It can obtain the performance near shannon limit in great amount of data transmission and memory channel; Be widely used in the various wireless communication systems; For example second generation satellite digital video broadcast standard (DVB-S2), China's military satellite communication system and global microwave H insert H interoperability technology (WiMax) etc., and it is up to a hundred million that the flank speed of supporting at present will reach, and will realize the data rate of hundreds of million future.
Why traditional LDPC sign indicating number is not widely used, and is because there are a plurality of defectives in it, and one of them defective is exactly that the coding techniques of LDPC sign indicating number is very complicated.If use the generator matrix of LDPC sign indicating number that information is encoded, need non-sparse check matrix of storage, and for the performance of pursuing the LDPC sign indicating number requires the code length long enough, this just causes this sparse check matrix can be very big, brings very big problem to storage.Following defective is arranged aspect design of encoder: one, the check matrix of traditional LDPC sign indicating number has stochastic behaviour; Need store a non-sparse check matrix during decoding; And require the code length long enough for the performance of pursuing the LDPC sign indicating number; This just causes this sparse check matrix can be very big, brings very big problem to storage; Two, have contradiction between the decoder throughput of traditional LDPC sign indicating number and the implementation complexity, if use serial decoding, then throughput must be very low, if use complete parallel or part parallel decoding, then the read-write of memory control must become quite loaded down with trivial details.
Summary of the invention
In order to overcome the deficiency of prior art, the object of the present invention is to provide the building method of the LDPC sign indicating number of simplicity of design.
The object of the invention is realized through following technical scheme:
The building method of LDPC sign indicating number may further comprise the steps:
(1) confirms the sparse check matrix H of first kind LDPC sign indicating number
1Structure, H
1Size is mP * nP, has following form:
H in the formula
I, jFor size is the matrix of P * P, the row of each matrix is heavy all to be fixed value 1 or 0, is obtained through cyclic shift by unit matrix I or null matrix; 1≤i≤m wherein, 1≤j≤n;
(2) select the degree of parallelism P and the parameter n of ldpc code decoder according to the demand of the communication system that is designed, then the code length of LDPC sign indicating number is nP, and wherein the value of parameter n is the common multiple of all code check denominators of communication system of being designed; According to P and n search parameter combination (w
Row, w
Col, k
2); W wherein
RowFor
Row heavy, w
ColFor
Column weight, k
2=(m
Wrow-3k)/(w
Col-3);
(3) according to the parameter combinations (w that searches
Row, w
Col, k
2), structure motherboard Matrix C, being defined as of said mother matrix Matrix C: use matrix H
I, jThe row mark of 1 position replaces in first row
In H
I, j, the matrix that uses the full null matrix of 0 replacement to obtain;
(4) according to the motherboard Matrix C of structure, use with the motherboard Matrix C in the corresponding submatrix H of each nonzero element
I, jSubstitute the nonzero element in the mother matrix Matrix C, use full null matrix to replace the neutral element in the mother matrix Matrix C, obtain
In conjunction with
Obtain the check matrix of first kind LDPC code word
The submatrix H that wherein nonzero element is corresponding with it
I, jCorresponding relation following: the value of nonzero element is submatrix H
I, jFirst row in the row mark of 1 position.
Can also construct the step of the method for the second type of LDPC sign indicating number that is associated with said first kind LDPC sign indicating number behind the completing steps (4):
(5) to H
1The row of matrix is pressed following rank order:
1,P+1,...,(m-1)P+1,2,P+2,...,(m-1)P+2,P,2P,...,mP;
1,P+1,...,(m-1)P+1,2,P+2,...,(m-1)P+2,P,2P,...,mP,
Promptly obtain the sparse check matrix H of second type of LDPC sign indicating number
2
The submatrix
of step (4) structure also satisfies following restrictive condition:
(a) the heavy w of row
RowIdentical, span is [3,40];
(b) value of column weight has two at most, and wherein less one is fixed value 3, and shared columns is designated as k
1P, another one is designated as w
Col, shared columns is designated as k
2P, wherein k
1+ k
2=k;
(c) value of column weight is according to arrange from big to small from left to right, corresponding to information bit from the highest significant bit to minimum effective bit;
Parameter combinations (the w that the said basis of step (3) searches
Row, w
Col, k
2), structure motherboard Matrix C specifically may further comprise the steps:
(3-1) design k
2Individual column weight is w
ColRow, be specially:
(3-1-1) produce w at random
ColNumbers different between individual [1, m] are listed as the rower that non-0 element belongs to, 1≤i≤k as i in the motherboard Matrix C
2If the heavy w that equaled of the row that rower is corresponding
Row, then will equal the heavy rower of row and be adjusted into the heavy minimum rower of row in nonoptional rower in the prostatitis;
(3-1-2) produce w at random
ColThe nonzero element that number between individual [1, P] is listed as as i in the motherboard Matrix C: said w
ColNumber satisfies: the value of (1) adjacent two row nonzero elements is different; (2) this w
ColThe value of any row corresponding row during any two values are listed as with the i-1 that produces before in the individual value is inequality;
(3-2) the design column weight is the k of 3 motherboard Matrix C
1=k-k
2Row are specially:
(3-2-1) produce numbers different between 3 [1, m] at random and be listed as the rower that non-0 element belongs to, k as i in the motherboard Matrix C
2+ 1≤i≤k; If the heavy w that equaled of the row that rower is corresponding
Row, then will equal the heavy rower of row and be adjusted into the heavy minimum rower of row in nonoptional rower in the prostatitis;
(3-2-2) produce the nonzero element of 3 numbers between [1, P] as i row among the C at random: said 3 numbers satisfy: the value of (1) adjacent two row nonzero elements is different; The value of any row corresponding row during (2) any two values are listed as with the i-1 that produces before in these 3 values is inequality;
The k that (3-3) step (3-1) is produced
2The k that row and step (3-2) produce
1The row amalgamation constitutes the motherboard Matrix C together.
The said search parameter combination of step (2) (w
Row, w
Col, k
2), detailed process is:
At w
Row(max (3,3k/m)≤w
Row≤40) and w
Col(3≤w
Col≤m) search in the scope, search and satisfy k
2=(mw
Row-3k)/(w
Col-3), and k
2Combination (w for nonnegative integer
Row, w
Col, k
2) be rational combination.
Completing steps (4) is also constructed the identical but process of the LDPC code check matrix that code length is different of the code check of the first kind LDPC sign indicating number that obtains with step (4) afterwards: use size as P ' * P ' (P '>P) matrix H "
I, jNonzero element in the mother matrix Matrix C that replacement step (3) obtains uses size to replace neutral element wherein as the full null matrix of P ' * P ', obtains matrix
Associate(d) matrix again
Obtain sparse check matrix
Compared with prior art, the present invention has the following advantages and technique effect:
(1) simplicity of design of the present invention only need simply be searched for just can construct and had uniform enconding complexity and the low first kind LDPC sign indicating number of required storage;
(2) sparse check matrix of the first kind LDPC sign indicating number of structure can be obtained the check matrix of second type of LDPC sign indicating number through the ordering of simple row and column; These two kinds of LDPC sign indicating number close relation; In practical application; If known a kind of yard coder then can simply sort to check digit at the input of the output of encoder and decoder and can obtain the coder of another kind of sign indicating number, greatly simplified design difficulty;
(3) in the first kind LDPC sign indicating number construction process, it is identical just to construct a plurality of code checks through simple substitution, and the first kind LDPC sign indicating number that code length is different has been simplified design difficulty greatly.
Description of drawings
Fig. 1 is the structure flow chart of two types of LDPC code check matrixes.
Fig. 2 is the cycle characteristics sketch map of the sparse check matrix of second type of LDPC sign indicating number.
Embodiment
Below in conjunction with embodiment and accompanying drawing, the present invention is done to specify further, but execution mode of the present invention is not limited thereto.
Embodiment
As shown in Figure 1, the building method of LDPC sign indicating number of the present invention may further comprise the steps:
(1) confirms the sparse check matrix H of first kind LDPC sign indicating number
1Structure, H
1Size is mP * nP, has following form:
H in the formula
I, jFor size is the matrix of P * P, the row of each matrix is heavy to be fixed value 1 or 0 all, is to be obtained through cyclic shift by unit matrix I or null matrix; 1≤i≤m wherein, 1≤j≤n;
The characteristics of
matrix have guaranteed that LDPC sign indicating number coding has linear complexity, and the characteristics of
matrix have guaranteed that LDPC code coding/decoding utensil has lower memory space.
(2) select the degree of parallelism P and the parameter n of ldpc code decoder according to the demand of the communication system that is designed, then the code length of LDPC sign indicating number is nP; Such as: when the data rate of design system is had relatively high expectations, should select bigger P; When requiring compatible various code rate, the n of selection should be the common multiple of all code check denominators, could guarantee that like this m and k are integer.For example, the compatible code check 1/3,1/2,2/3,3/4 of system requirements, 4/5,7/8 o'clock, then n should be all code check denominators 3,2,3,4,5, the integral multiple of the least common multiple 120 of 8}.
(b)
The value of column weight have two at most, wherein less one is fixed value 3, shared columns is designated as k
1P, another one is designated as w
Col, shared columns is designated as k
2P, wherein k
1+ k
2=k;
(c) value of column weight arrives minimum effective bit (LSB, Least Significant Bits) corresponding to information bit from the highest significant bit (MSB, Most Significant Bits) according to arranging from left to right from big to small;
(d) with
The check matrix H that the combination back is formed
1Ring length is not 4 ring (being called for short ring 4), can resolve into two conditions: (1)
Do not encircle 4, (2)
Any row in any two adjacent non-zero submatrices H
I, jAnd H
I+1, jInequality.According to
Characteristics can know, like this
With
H after the combination
1Just can not constitute ring length and be 4 ring.
Search procedure is specially:
Know w according to condition (b)
ColThe hunting zone be 3≤w
Col≤m, and by matrix
In 1 number is fixing can know:
(3k
1+w
colk
2)P=mPw
row(1)
Again because w
Col>=3, k
1+ k
2=k, therefore:
k
2=(mw
row-3k)/(w
col-3)(2)
w
row≥3k/m (3)
To sum up can know w
RowThe hunting zone be max (3,3k/m)≤w
Row≤40.
To w
Row(max (3,3k/m)≤w
Row≤40) and w
Col(3≤w
Col≤m) in the scope of appointment, search for, search and satisfy formula (2) and be the combination (w of nonnegative integer
Row, w
Col, k
2) be exactly reasonably combination.
(3) according to the parameter combinations (w that searches
Row, w
Col, k
2), structure motherboard Matrix C, being defined as of said mother matrix Matrix C: use matrix H
I, jThe row mark of 1 position substitutes in first row
In H
I, jThe matrix that (full null matrix replaces with 0) obtains, for example, matrix H
I, jFirst row in 1 be positioned at the 3rd row, then use 3 to replace
In H
I, j
Specifically may further comprise the steps of structure motherboard Matrix C:
(3-1) design k
2Individual column weight is w
ColRow, be specially:
(3-1-1) produce w at random
ColNumbers different between individual [1, m] are listed as the rower that non-0 element belongs to, 1≤i≤k as i among the C
2If the heavy w that equaled of the row that rower is corresponding
Row, then will equal the heavy rower of row and be adjusted into the heavy minimum rower of row in nonoptional rower in the prostatitis;
(3-1-2) produce w at random
ColThe nonzero element that number between individual [1, P] is listed as as i among the C: said w
ColNumber satisfies: the value of (1) adjacent two row nonzero elements is different; (2) this w
ColThe value of any row corresponding row during any two values are listed as with the i-1 that produces before in the individual value is inequality; If do not satisfy wherein one, then need adjust this w
ColThe position of number.If through the value that does not still satisfy condition after the several times adjustment, then need restart the search procedure of (3-1-2).Still do not find suitable value if pass through several times (3-1-2) process again, then return (3-1-1) and search for again till satisfying condition;
(3-2) the design column weight is the k of 3 C
1=k-k
2Row are specially:
(3-2-1) produce numbers different between 3 [1, m] at random and be listed as the rower that non-0 element belongs to, k as i among the C
2+ 2≤i≤k; If the heavy w that equaled of the row that rower is corresponding
Row, then will equal the heavy rower of row and be adjusted into the heavy minimum rower of row in nonoptional rower in the prostatitis;
(3-2-2) produce the nonzero element of 3 numbers between [1, P] as i row among the C at random: said 3 numbers satisfy: the value of (1) adjacent two row nonzero elements is different; The value of any row corresponding row during (2) any two values are listed as with the i-1 that produces before in these 3 values is inequality;
The k that (3-3) step (3-1) is produced
2The k that row and step (3-2) produce
1The row amalgamation constitutes the motherboard Matrix C together.
(4) according to the motherboard Matrix C of structure, use with the motherboard Matrix C in the corresponding submatrix H of each element value
I, jThe element (0 element uses full null matrix to replace) that substitutes in the mother matrix Matrix C obtains
And then obtain
The submatrix H that wherein nonzero element is corresponding with it
I, jCorresponding relation following: the value of nonzero element is submatrix H
I, jFirst row in the row mark of 1 position, for example nonzero element is 3, then its corresponding submatrix H
I, jFirst the row in 1 be positioned at the 3rd row.
Can be known by top search condition, if the motherboard Matrix C satisfies the search condition that parameter is P, then necessarily satisfy the structural environment of the motherboard matrix of P '>P, be the submatrix H of P ' * P ' with size therefore
I, jElement in the replaced C can obtain the submatrix that parameter is P ' time
The LDPC sign indicating number code check that construct this moment is identical with code word code check before, and just code length is different, and detailed process is:
Use the matrix H of size " as P ' * P '
I, jNonzero element in the mother matrix Matrix C that replacement step (3) obtains uses size to replace neutral element wherein as the full null matrix of P ' * P ', obtains matrix
Associate(d) matrix again
Obtain sparse check matrix
According to H
1" encode;
H wherein "
I, jFor size is the matrix of P ' * P ', P '>P, the row of each matrix is heavy to be fixed value 1 or 0 all, is to be obtained through cyclic shift by unit matrix I or null matrix.
If input information bits s=is [s
1, s
2, L, s
KP], the check bit p=[p that calculates
1, p
2, L, p
MP], then the coding method of first kind LDPC sign indicating number is following:
Step 1: utilize input information vector s, calculate intermediate object program x=[x
1, x
2, L, x
MP]:
Step 2: utilize x computation of parity bits p, adopt following formula:
At first obtain first check bit p during coding
1=x
1, obtain successively then
Code word c=[c behind the coding
1, c
2, L, c
NP]=[s|p].
Construct the sparse check matrix H of the second type of LDPC sign indicating number that is associated with first kind LDPC sign indicating number if desired
2, then can carry out through following steps:
To H
1The row of matrix is pressed following rank order:
1,P+1,...,(m-1) P+ 1,2,P+2,...,(m-1)P+ 2,P,2P,...,mP;
1,P+1,...,(m-1) P+ 1,2,P+2,...,(m-1)P+ 2,P,2P,...,mP
Can obtain the sparse check matrix H of second type of LDPC sign indicating number
2
Second type of LDPC sign indicating number is that non-canonical repeats accumulation sign indicating number (IRA), the sparse check matrix H of such LDPC sign indicating number
2Following characteristic is arranged:
The characteristics of
matrix have guaranteed that LDPC sign indicating number coding has linear complexity, and the characteristics of
matrix have guaranteed that LDPC code coding/decoding utensil has lower memory space.
Each A wherein
i(1≤i≤k) inner column weight is all identical, and size is mP * P, and is circular matrix, has cycle characteristics shown in Figure 2, and the black box that has same numeral among the figure representes that the element on the check matrix correspondence position is 1, and colourless square corresponds to 0.Can find out by Fig. 2; The row at the square of same numeral place number can be by the row number decision of the square of same label in any row; Therefore the square row difference of two row same numerals that face mutually is fixed as m=2mod (mP) (mod is modulo operation, wherein P=16), each A in this case
iOnly need in storage one row 1 row number, then the row through calculating other row number.Generally storing the row number at 1 place in the 1st row, also is initial verification address.
The coding of second type of LDPC sign indicating number (being the IRA sign indicating number) divides following two steps to carry out:
Step 1: utilize input information vector s, calculate intermediate object program x=[x
1, x
2, L, x
MP]:
Step 2: utilize x computation of parity bits p, adopt following formula:
At first obtain first check bit p during coding
1=x
1, obtain successively then
Code word c=[c behind the coding
1, c
2, L, c
NP]=[s|p].
Discuss the relation between two types of LDPC sign indicating numbers below.
At first contrast the difference part of both check matrixes:
The code word c and the relation between the H that are obtained by the check matrix H coding are:
cH
T=0(7)
Can know by formula (7), the row of check matrix H is exchanged the result who does not influence formula (6), just not influence coding result.We can be at first with H thus
1Row adjust (matrix after the note adjustment is:
Matrix after the feasible adjustment
Part satisfies second type of LDPC sign indicating number
Form, just each size all need satisfy characteristic of circular matrix for the matrix in block form of mP * P.For this reason, we discuss H
1In the matrix by H
1, i, H
2, i..., H
M, iThe size that submatrix is formed is mP * P matrix A
i(1≤i≤k) wherein:
With A
iRow resequence, sort method is: extract each submatrix H
1, i, H
2, i..., H
M, iJ capablely form a matrix in order and (be designated as B
j, 1≤j≤P), and then with B
jForm in order matrix A '
i:
Because submatrix H
I, jBe to be obtained through cyclic shift by unit matrix I, have characteristic: a is listed as the mobile element that circulates downwards just can obtain a+1 row (if 1 place behavior last column that a is listed as, then next behavior first is gone, referring to following formula).
Characteristic is found out A ' easily thus
iThe circulation downwards of a row move m element and just can obtain a+1 and be listed as.That is to say A '
iIt is circular matrix.Because all A '
i(wherein 1≤i≤k) obtains circular matrix so matrix through identical row adjustment
Satisfy second type of LDPC sign indicating number
Characteristic, wherein:
Must, with matrix H
1Row arrange again according to following order, can make the adjustment after matrix H
1'
Part satisfies second type of LDPC sign indicating number
Form:
1,P+1,...,(m-1) P+ 1,2,P+2,...,(m-1)P+ 2,P,2P,...,mP
(8)
Can know by top analysis, with H
1Be converted into the matrix H of equivalence through line ordering
1' afterwards, H
1With H
2Difference only be present in
With
Part.
Analyze H
1Cataloged procedure:
At first obtain first check bit p during coding
1=x
1, obtain successively then
If:
Can know x by formula (6)
iWith
Row corresponding one by one, according to
With
Corresponding relation can know that the corresponding relation of x and x ' is:
{x
1,x
P+1,...,x
(m-1)P+1,x
2,x
P+2,...,x
(m-1)P+2,x
2P,...,x
mP}={x′
1,x′
2,x′
3,...,x′
mP}(10)
Hence one can see that, and cataloged procedure becomes:
Order
{p
1,p
P+1,...,p
(m-1)P+1,p
2,p
P+2,...,p
(m-1)P+2,p
P,p
2P,...,p
mP}={p′
1,p′
2,p′
3,...,p′
mP}(11)
Then cataloged procedure becomes:
This process and H
2Cataloged procedure identical.Can know H from top derivation
1With H
2Difference between the code word is that the check digit order is different, and that the order of check digit correspondence is H
pThe order of rectangular array, hence one can see that, will
Row arrange H according to the order shown in the formula (7)
1' can satisfy H
2The characteristics of matrix.
Because just there is difference in two types of LDPC sign indicating numbers of being constructed on the order of check digit; If therefore the coder of known one type of LDPC sign indicating number will be constructed the coder of another kind of LDPC sign indicating number; Only need sort to the input of the output of encoder and decoder gets final product, and does not need extra modification.
Known H
1, obtain H after the ordering through row and column
2Process be not reversible.Suppose known H
2, the matrix that obtains after sorting according to top inverse process is H
2', then
H ' wherein
I, jBe that size is the matrix of P * P, the row of each matrix is heavy all to be fixed value (being not limited to 1 and 0), when row heavily is r
wThe time, H '
I, jBe by r
wThe unit matrix I of individual different process cyclic shift adds up and obtains.
The foregoing description is a preferred implementation of the present invention; But execution mode of the present invention is not limited by the examples; Other any do not deviate from change, the modification done under spirit of the present invention and the principle, substitutes, combination, simplify; All should be the substitute mode of equivalence, be included within protection scope of the present invention.
Claims (6)
1.LDPC the building method of sign indicating number is characterized in that, may further comprise the steps:
(1) confirms the sparse check matrix H of first kind LDPC sign indicating number
1Structure, H
1Size is mP * nP, has following form:
H in the formula
I, jFor size is the matrix of P * P, the row of each matrix is heavy all to be fixed value 1 or 0, is obtained through cyclic shift by unit matrix I or null matrix; 1≤i≤m wherein, 1≤j≤n;
Wherein the matrix represented of
upper right corner a has following structure:
(2) select the degree of parallelism P and the parameter n of ldpc code decoder according to the demand of the communication system that is designed, then the code length of LDPC sign indicating number is nP, and wherein the value of parameter n is the common multiple of all code check denominators of communication system of being designed; According to P and n search parameter combination (w
Row, w
Col, k
2); W wherein
RowFor
Row heavy, w
ColFor
Column weight, k
2=(mw
Row-3k)/(w
Col-3);
(3) according to the parameter combinations (w that searches
Row, w
Col, k
2), structure motherboard Matrix C, being defined as of said mother matrix Matrix C: use matrix H
I, jThe row mark of 1 position replaces in first row
In H
I, j, the matrix that uses the full null matrix of 0 replacement to obtain;
(4) according to the motherboard Matrix C of structure, use with the motherboard Matrix C in the corresponding submatrix H of each nonzero element
I, jSubstitute the nonzero element in the mother matrix Matrix C, use full null matrix to replace the neutral element in the mother matrix Matrix C, obtain
In conjunction with
Obtain the check matrix of first kind LDPC code word
The submatrix H that wherein nonzero element is corresponding with it
I, jCorresponding relation following: the value of nonzero element is submatrix H
I, jFirst row in the row mark of 1 position.
2. the building method of LDPC sign indicating number according to claim 1 is characterized in that, completing steps
(4) also construct the step of the method for the second type of LDPC sign indicating number that is associated with said first kind LDPC sign indicating number after:
(5) to H
1The row of matrix is pressed following rank order:
1,P+1,...,(m-1)P+1,2,P+2,...,(m-1)P+2,P,2P,...,mP;
Then the row of submatrix
are arranged according to following order
1, P+1 ..., (m-1) P+1,2, P+2 ..., (m-1) P+2, P, 2P ..., mP obtains the sparse check matrix H of second type of LDPC sign indicating number
2
3. the building method of LDPC sign indicating number according to claim 1; It is characterized in that the submatrix
of step (4) structure satisfies following restrictive condition:
(a) the heavy w of row
RowIdentical, span is [3,40];
(b) value of column weight has two at most, and wherein less one is fixed value 3, and shared columns is designated as k
1P, another one is designated as w
Col, shared columns is designated as k
2P, wherein k
1+ k
2=k;
(c) value of column weight is according to arrange from big to small from left to right, corresponding to information bit from the highest significant bit to minimum effective bit;
(d) with
The check matrix H of forming
1Ring length is not 4 ring.
4. the building method of LDPC sign indicating number according to claim 3 is characterized in that, the parameter combinations (w that the said basis of step (3) searches
Row, w
Col, k
2), structure motherboard Matrix C specifically may further comprise the steps:
(3-1) design k
2Individual column weight is w
ColRow, be specially:
(3-1-1) produce w at random
ColNumbers different between individual [1, m] are listed as the rower that non-0 element belongs to, 1≤i≤k as i in the motherboard Matrix C
2If the heavy w that equaled of the row that rower is corresponding
Row, then will equal the heavy rower of row and be adjusted into the heavy minimum rower of row in nonoptional rower in the prostatitis;
(3-1-2) produce w at random
ColThe nonzero element that number between individual [1, P] is listed as as i in the motherboard Matrix C: said w
ColNumber satisfies: the value of (1) adjacent two row nonzero elements is different; (2) this w
ColThe value of any row corresponding row during any two values are listed as with the i-1 that produces before in the individual value is inequality;
(3-2) the design column weight is the k of 3 motherboard Matrix C
1=k-k
2Row are specially:
(3-2-1) produce numbers different between 3 [1, m] at random and be listed as the rower that non-0 element belongs to, k as i in the motherboard Matrix C
2+ 1≤i≤k; If the heavy w that equaled of the row that rower is corresponding
Row, then will equal the heavy rower of row and be adjusted into the heavy minimum rower of row in nonoptional rower in the prostatitis;
(3-2-2) produce the nonzero element of 3 numbers between [1, P] as i row in the motherboard Matrix C at random: said 3 numbers satisfy: the value of (1) adjacent two row nonzero elements is different; The value of any row corresponding row during (2) any two values are listed as with the i-1 that produces before in these 3 values is inequality;
The k that (3-3) step (3-1) is produced
2The k that row and step (3-2) produce
1The row amalgamation constitutes the motherboard Matrix C together.
5. the building method of LDPC sign indicating number according to claim 3 is characterized in that, the said search parameter combination of step (2) (w
Row, w
Col, k
2), detailed process is:
At w
Row(max (3,3k/m)≤w
Row≤40) and w
Col(3≤w
Col≤m) search in the scope, search and satisfy k
2=(mw
Row-3k)/(w
Col-3), and k
2Combination (w for nonnegative integer
Row, w
Col, k
2) be rational combination.
6. the building method of LDPC sign indicating number according to claim 1; It is characterized in that completing steps (4) is also constructed the identical but process of the LDPC code check matrix that code length is different of the code check of the first kind LDPC sign indicating number that obtains with step (4) afterwards: use size as P ' * P ' (P '>P) matrix H "
I, jNonzero element in the mother matrix Matrix C that replacement step (3) obtains uses size to replace neutral element wherein as the full null matrix of P ' * P ', obtains matrix
Associate(d) matrix again
Obtain sparse check matrix
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2011103414024A CN102386932B (en) | 2011-11-02 | 2011-11-02 | LDPC code constitution method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2011103414024A CN102386932B (en) | 2011-11-02 | 2011-11-02 | LDPC code constitution method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102386932A true CN102386932A (en) | 2012-03-21 |
CN102386932B CN102386932B (en) | 2013-11-13 |
Family
ID=45825951
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2011103414024A Active CN102386932B (en) | 2011-11-02 | 2011-11-02 | LDPC code constitution method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102386932B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018059588A1 (en) * | 2016-09-30 | 2018-04-05 | 中兴通讯股份有限公司 | Quasi-cyclic ldpc coding and decoding method and apparatus, and ldpc coder and decoder |
CN112821895A (en) * | 2021-04-16 | 2021-05-18 | 成都戎星科技有限公司 | Code identification method for realizing high error rate of signal |
CN117040543A (en) * | 2023-10-09 | 2023-11-10 | 苏州元脑智能科技有限公司 | Error correction coding check matrix generation method, device and storage medium |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101047387A (en) * | 2007-03-23 | 2007-10-03 | 北京大学 | Construction method of multi-code rate compatible LDPC code and its decoder |
CN101141133A (en) * | 2007-10-23 | 2008-03-12 | 北京邮电大学 | Method of encoding structured low density check code |
-
2011
- 2011-11-02 CN CN2011103414024A patent/CN102386932B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101047387A (en) * | 2007-03-23 | 2007-10-03 | 北京大学 | Construction method of multi-code rate compatible LDPC code and its decoder |
CN101141133A (en) * | 2007-10-23 | 2008-03-12 | 北京邮电大学 | Method of encoding structured low density check code |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018059588A1 (en) * | 2016-09-30 | 2018-04-05 | 中兴通讯股份有限公司 | Quasi-cyclic ldpc coding and decoding method and apparatus, and ldpc coder and decoder |
CN107888198A (en) * | 2016-09-30 | 2018-04-06 | 中兴通讯股份有限公司 | Quasi-cyclic LDPC coding and decoding method, device and LDPC coders |
US11133826B2 (en) | 2016-09-30 | 2021-09-28 | Zte Corporation | Quasi-cyclic LDPC coding and decoding method and apparatus, and LDPC coder and decoder |
US11637568B2 (en) | 2016-09-30 | 2023-04-25 | Zte Corporation | Quasi-cyclic LDPC coding and decoding method and apparatus, and LDPC coder and decoder |
CN107888198B (en) * | 2016-09-30 | 2023-05-26 | 中兴通讯股份有限公司 | Quasi-cyclic LDPC (low density parity check) coding and decoding method and device and LDPC coder and decoder |
CN112821895A (en) * | 2021-04-16 | 2021-05-18 | 成都戎星科技有限公司 | Code identification method for realizing high error rate of signal |
CN117040543A (en) * | 2023-10-09 | 2023-11-10 | 苏州元脑智能科技有限公司 | Error correction coding check matrix generation method, device and storage medium |
CN117040543B (en) * | 2023-10-09 | 2024-02-20 | 苏州元脑智能科技有限公司 | Error correction coding check matrix generation method, device and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN102386932B (en) | 2013-11-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR102014905B1 (en) | Transmitting apparatus and interleaving method thereof | |
CN100596029C (en) | Method of constructing check matrix for LDPC code, and encoding and decoding device of using the method | |
KR101986778B1 (en) | Transmitting apparatus and interleaving method thereof | |
CN104868925B (en) | Coding method, interpretation method, code device and the code translator of structured LDPC code | |
US10536169B2 (en) | Encoder and decoder for LDPC code | |
CN101741396B (en) | Method and device for coding or decoding low-density parity-check (LDPC) codes with variable code length, coder and decoder | |
CN101032082B (en) | Method and apparatus for encoding and decoding data | |
CN101395804B (en) | Method of encoding and decoding using ldpc code | |
KR102018342B1 (en) | Transmitting apparatus and interleaving method thereof | |
CN101162907B (en) | Method and device for constructing low-density parity code check matrix | |
CN101164241B (en) | Encoding apparatus and encoding method | |
JP4822452B2 (en) | Encoding or decoding method and apparatus using LDPC code | |
KR101965371B1 (en) | Transmitting apparatus and interleaving method thereof | |
CN101567697B (en) | Coder and method for coding rate-compatible low-density parity-check codes | |
CN103069720B (en) | Error correction decoding device and error correction decoding method | |
CN102468855A (en) | Apparatus and method for correcting at least one bit error within a coded bit sequence | |
US9104589B1 (en) | Decoding vectors encoded with a linear block forward error correction code having a parity check matrix with multiple distinct pattern regions | |
WO2012159304A1 (en) | Encoding method and apparatus for low density parity check code | |
US8145971B2 (en) | Data processing systems and methods for processing digital data with low density parity check matrix | |
JP2014523168A (en) | Method and apparatus for transmitting data | |
US20080320374A1 (en) | Method and apparatus for decoding a ldpc code | |
CN101800627B (en) | Hardware implementation of multiple-code-rate-compatible high-speed low-density parity-check code encoder | |
CN102386932B (en) | LDPC code constitution method | |
CN102412844A (en) | Decoding method and decoding device of IRA (irregular repeat-accumulate) series LDPC (low density parity check) codes | |
CN102611465B (en) | Coder of structured q-ary irregular repeat-accumulate code and coding method of coder |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |