CN102412842A - Method and device for encoding low-density parity check code - Google Patents

Abstract

The invention provides a method and device for encoding a low-density parity check code. The method comprises the following steps of: comparing a mother code rate and a transmission code rate which are supported by a uniform basic matrix; determining the length K'of an input information grouping bit according to a comparison result; determining an encoding matrix of the low-density parity check (LDPC) code according to the uniform basic matrix and an expansion factor; determining an information bit to be encoded according to the received information grouping bit with the length of K'; and performing LDPC encoding on the information bit to be encoded by using the encoding matrix, and processing a mother code word bit obtained through encoding to acquire a code word bit to be transmitted. By the method and device for encoding the LDPC code, the basic matrix and the expansion factor are determined; the input information grouping length is determined or the encoding matrix is processed according to the received transmission code rate, so the code word which is in accordance with the transmission code rate can be acquired; on the basis of the optimal performance, the changeability of the code rate is realized; furthermore, the hardware design complexity is reduced and the cost is lower.

Description

A kind of coding method of low density parity check code and device

Technical field

The present invention relates to digital communicating field, relate in particular to a kind of coding method and device of low density parity check code.

Background technology

At present, digital communication system is the communication system of using always.As shown in Figure 1, digital communication system is made up of transmitting terminal, channel and receiving terminal, and wherein, transmitting terminal generally includes information source, source encoder, channel encoder and modulator parts such as (or writing units); Receiving terminal generally includes demodulator (or sensing element), channel decoder, source decoder and the stay of two nights; There is channel (or storage medium) between transmitting terminal and the receiving terminal; And exist noise source in the channel in digital communication system; And channel code chain (comprising channel decoding, modulation etc.) is the most critical technology of whole digital communication physical layer, and it has determined the validity and the reliability of the transmission of digital communication system bottom.

Channel encoder is in order to resist various noises and interference in the transmission course, and through increasing redundant information artificially, the system that makes has the ability of automatic correction mistake, thus the reliability of guarantee information transmission.Low-density checksum (Low Density Parity Check; LDPC) sign indicating number is the one type of linear block codes that can use very sparse parity matrix or bipartite graph definition; Utilize the sparse property of its check matrix just; Could realize the coding and decoding of low complex degree, thereby make LDPC move towards practicability.

The parity check matrix H of a M * N has defined the constraint that each code word with N bit satisfies M parity check collection, in the N bit codewords that promptly coding obtains, has the check digit of M bit, an information bit of N-M bit.If the parity check matrix H of LDPC sign indicating number is (m _b* z) * (n _b* z) matrix, it is by m _b* n _bIndividual matrix in block form constitutes, and each matrix in block form all is the different powers of the basic permutation matrix of z * z, and when basic permutation matrix was unit matrix, they all were the cyclic shift matrices of unit matrix, as move to right, and have following form:

$H=\left[\begin{array}{ccccc}{P}^{h_{00}^b}& P^{h_{01}^b}& P^{h_{02}^b}& \·\·\·& P^{h_{0n_{b-1}}^b}\\ P^{h_{10}^b}& P^{h_{11}^b}& P^{h_{12}^b}& \·\·\·& P^{h_{1n_{b-1}}^b}\\ \·\·\·& \·\·\·& \·\·\·& \·\·\·& \·\·\·\\ P^{h_{\mathrm{mb}-10}^b}& P^{h_{m_b-11}^b}& P^{h_{m_b-12}^b}& \·\·\·& P^{h_{m_b-1n_b-1}^b}\end{array}\right]=P^{H_b}$

Just can each matrix in block form of unique identification through such power time

; Time available 0 expression of the power of unit matrix, the power of null matrix is represented with-1.Like this, if all use its power to replace each matrix in block form of H, just obtain a m _b* n _bPower submatrix H _bHere define H _bBe the basis matrix of H, H is called H _bExtended matrix.When actual coding, the columns=N/n of z=code length/basis matrix _b, be called spreading factor.For example, matrix

Can use following parameter z, one 2 * 4 basis matrix H _bThe i.e. unit matrix expansion of a z * z obtains with basic permutation matrix,

Z=3 with $H_b=\left[\begin{array}{cccc}0& 1& 0& -1\\ 2& 1& 2& 1\end{array}\right]$

Therefore, encoder that also we can say the LDPC sign indicating number is by basis matrix H _b, spreading factor z and the unique generation of basic permutation matrix.

If for each different spreading factor, LDPC adopts a basis matrix, so for each different code length; The coder of LDPC sign indicating number all need be stored a basis matrix; When code length is a variety of, just need a lot of basis matrixs of storage, consider problems such as storage with regard to needs.Therefore; When needing to realize becoming code length; The LDPC sign indicating number of multiple code length can use the basis matrix of same form in the certain limit of same code check; When we call unified basis matrix

different code length to this matrix;

revised and expand; Can obtain parity check matrix H, make the coder that generates go for the variable occasion of code length.

Wherein, correction is to utilize the spreading factor of different code length to basis matrix H _bOr unified basis matrix

In nonnegative value revise, make to revise the back element value less than the spreading factor value under this code length.

This is because H _bOr

In element represented the power time of basic permutation matrix, and basic permutation matrix, the i.e. unit matrix of z * z; Every circulation primary; Be equivalent to one of unit matrix ring shift right, i.e. unit matrix ring shift right one row, so the maximum cyclic shifts of basic circular matrix z-1 time; If circulation z time, the still unit matrix that obtains is equal to nothing and carries out cyclic shift.Correction algorithm has a variety of, for example, can adopt delivery (mod), round (scale+floor) or round off (scale+round) etc., establishes P _IjBe basis matrix H _bIn the non-negative element of the capable j of i row, P ' _IjBe revised basis matrix

In the non-negative element of the capable j of i row, have:

For delivery (mod) algorithm: $P_{\mathrm{Ij}}^{\′}\≡P_{\mathrm{Ij}}\mathrm{Mod}z\≡P_{\mathrm{Ij}}\mathrm{Mod}\frac{N}{n_b};$

For rounding (scale+floor) algorithm:

For (scale+round) algorithm that rounds off: $P_{\mathrm{Ij}}^{\′}=\mathrm{Round}(P_{\mathrm{Ij}}\×\frac{z}{z_{\mathrm{Max}}})=\mathrm{Round}(P_{\mathrm{Ij}}\×\frac{N}{N_{\mathrm{Max}}})$

Wherein, z is the corresponding spreading factor of current code length, i.e. the line number of piecemeal square formation or columns, z _MaxFor maximum is supported code length N _MaxCorresponding spreading factor.Mod is a modulo operation;

rounds operation under being, Round is the operation that rounds up.

For example, for the LDPC sign indicating number of code length 1152 bits, establish its basis matrix H _bCertain non-negative element be 93, the maximum code length of establishing its support is 2304, the size of basis matrix is 12 * 24, it is carried out correction result be:

For delivery (mod) algorithm: $93\mathrm{Mod}\frac{1152}{24}=93\mathrm{Mod}48=45;$

For rounding (scale+floor) algorithm:

For (scale+round) algorithm that rounds off: $\mathrm{Round}(93\×\frac{1152}{2304})=\mathrm{Round}\left(46.5\right)=47.$

The LDPC sign indicating number of this specific code check variable code length is because the basis matrix with same form; So can use an encoder/decoder fully, wherein the basis matrix of its same form also can be called unified basis matrix

The size of supposing the parity check matrix H of a LDPC sign indicating number is (m _b* z) * (n _b* z), unified basis matrix

Size be m _b* n _b, female sign indicating number code check R of supporting of this LDPC sign indicating number then _m=(n _b-m _b)/n _bIf the information block length of input is K ', obtain the code word bits to be transmitted that length is N ' through after a series of processing, then claim transmission code rate R _Tx=K '/N '.

In IEEE802.16e, the code check of LDPC sign indicating number support is 1/2,2/3,3/4,5/6, in the encoding scheme of existing LDPC sign indicating number; Usually adopt the LDPC encoder matrix H of different female sign indicating number code checks; To guarantee that the LDPC sign indicating number has to a certain degree code check and performance flexibly, at this moment, four basis matrix H will occur _b, on hardware, need a plurality of LDPC coders, can roll up the realization cost of hardware; Perhaps realize the coding and decoding of different code checks, but hardware designs can be very complicated, can roll up hard-wired cost equally with an encoder/decoder that can adapt to multiple female sign indicating number code check.

Summary of the invention

In view of this, main purpose of the present invention is to provide a kind of coding method and device of LDPC sign indicating number, has realized the coding of the LDPC sign indicating number of any variable bit rate.

For achieving the above object, technical scheme of the present invention is achieved in that

A kind of coding method of low-density checksum LDPC sign indicating number, said method comprises the steps:

The female sign indicating number code check that relatively more unified basis matrix is supported and the size of transmission code rate based on comparative result, are confirmed the length K of input information block bit ';

Confirm the encoder matrix of LDPC sign indicating number according to unifying basis matrix and spreading factor;

Length by receiving is the information block bit of K ', confirms information bit to be encoded;

Utilize said encoder matrix to treat the information encoded bit and carry out the LDPC coding, and female sign indicating number code word bits that coding obtains is handled, obtain code word bits waiting for transmission.

Said method also comprises: unified basis matrix and spreading factor that the LDPC sign indicating number is set.

Also comprise before the female sign indicating number code check of said relatively more unified basis matrix support and the size of transmission code rate: receive current transmission code rate;

Said according to comparative result, confirm the length K of input information packet bit ' be:

Work as R _Tx>=R _mThe time, confirm the length K of information block ' be (n _b-m _b) * z; Work as R _Tx＜R _mThe time, the length of confirming the information block bit is K '=func (m _b* z * R _Tx/ (1-R _Tx)) or K '=func (m _b* z/ (1-R _Tx))-m _b* z or K '=func (func (m _b* z/ (1-R _Tx)) * R _Tx);

Wherein, R _TxBe transmission code rate, R _mBe female sign indicating number code check, m _b, n _bBe respectively the line number and the columns of unified basis matrix, z is a spreading factor, and func (x) is a kind of of number, round numbers downwards or the round numbers of rounding off that x is rounded up.

Said based on unified basis matrix and spreading factor, confirm that the encoder matrix of LDPC sign indicating number is:

Unified basis matrix is revised,, utilized the unit matrix of z * z that the basis matrix that correction obtains is expanded, obtain the encoder matrix of LDPC sign indicating number according to spreading factor z.

Said is the information block bit of K ' by the length that receives, and confirms that information bit to be encoded is:

Work as R _Tx>=R _mThe time, be that the information block bit of K ' is directly as information bit to be encoded with the length that receives; Work as R _Tx＜R _mThe time,, length adds filling bit in being the information block bit of K ', and formation length is k _bThe information bit to be encoded of * z.

Said female sign indicating number code word bits that coding is obtained is treated to:

Female sign indicating number code word to coding obtains is reset, and the female sign indicating number code word bits before wherein resetting is A ₀, A ₁..., A _{Nb * z-1}, the female sign indicating number code word bits after the rearrangement is B ₀, B ₁..., B _{Nb * z-1}Rearrangement formulae is:

Wherein, k _b=n _b-m _b, PV representes code word rearrangement vector, its element set is { n _b-m _b, n _b-m _b+ 1, n _b-m _b+ 2 ..., n _b-1}, index is the element of x in PV (x) the expression PV vector, x is a nonnegative integer;

Work as R _Tx>=R _mThe time, the female sign indicating number code word bits B after resetting ₀, B ₁..., B _{Nb * z-1}According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission;

Work as R _Tx＜R _mThe time, with the said k that adds _bThe female sign indicating number code word bits B of the individual filling bit of * z-K ' after reset ₀, B ₁, B _{Nb * z-1}In systematic bits in delete, remaining female sign indicating number code word bits is as code word bits waiting for transmission.

The encoder matrix of said definite LDPC sign indicating number is:

Unified basis matrix is revised,, utilized the unit matrix of z * z that the basis matrix that correction obtains is expanded, obtain the extended matrix of LDPC sign indicating number according to spreading factor z;

Work as R _Tx>=R _mThe time, with the encoder matrix of extended matrix as the LDPC sign indicating number; Work as R _Tx＜R _mThe time, the func (m on intercepting extended matrix the right _b* z/ (1-R _Tx)) row, the matrix that intercepting is obtained is as the encoder matrix of LDPC sign indicating number; Perhaps,

As

when being positive integer; Unified basis matrix is revised, obtained revised basis matrix;

Work as R _Tx>=R _mThe time, according to spreading factor z, utilize the unit matrix of z * z that the basis matrix that correction obtains is expanded, obtain the encoder matrix of LDPC sign indicating number; Work as R _Tx＜R _mThe time, the k ' on the revised basis matrix of intercepting the right _b+ m _bRow according to spreading factor z, utilize the unit matrix of z * z that the matrix that intercepting obtains is expanded, and obtain the encoder matrix of LDPC sign indicating number.

Said is the information block bit of K ' by the length that receives, and confirms that information bit to be encoded is: with length is that the information block bit of K ' is directly as information bit to be encoded.

Said female sign indicating number code word bits that coding is obtained is treated to:

Female sign indicating number code word to coding obtains is reset, and the female sign indicating number code word bits after the female sign indicating number code word bits before wherein resetting is reset for

for

rearrangement formulae is:

wherein;

PV representes code word rearrangement vector; Its element set is the element of x for index in

PV (x) expression PV vector; X is a nonnegative integer, and wherein

expression is to the x number that rounds up;

Work as R _Tx>=R _mThe time, the female sign indicating number code word bits after resetting According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission;

Work as R _Tx＜R _mThe time, directly will be from the female sign indicating number code word bits after resetting

As code word bits waiting for transmission.

A kind of code device of LDPC sign indicating number comprises coding unit, and said device also comprises: comparing unit, packet bit confirm that unit, encoder matrix confirm that unit, information bit confirm unit and code word processing unit; Wherein,

Comparing unit is used for the female sign indicating number code check of relatively more unified basis matrix support and the size of transmission code rate;

Packet bit is confirmed the unit, is used for the comparative result according to comparing unit, confirms the length K of input information packet bit ';

Encoder matrix is confirmed the unit, is used for confirming the encoder matrix of LDPC sign indicating number based on unifying basis matrix and spreading factor;

Information bit is confirmed the unit, is used for receiving obtaining the information block bit that length is K ', and confirms information bit to be encoded, and it is transferred to coding unit;

The code word processing unit is used for that coding unit is carried out the female sign indicating number code word that LDPC coding obtains and handles, and obtains code word bits waiting for transmission.

Said device also comprises: the unit is set, is used to be provided with the unified basis matrix and the spreading factor of LDPC sign indicating number.

Said device also comprises input unit, is used to receive current transmission code rate;

Said packet bit confirms that the comparative result that the unit specifically is used for when comparing unit is R _Tx>=R _mThe time, confirm the length K of information block bit ' be (n _b-m _b) * z; When the comparative result of comparing unit is R _Tx＜R _mThe time, the length of confirming the information block bit is K '=func (m _b* z * R _Tx/ (1-R _Tx)) or K '=func (m _b* z/ (1-R _Tx))-m _b* z or K '=func (func (m _b* z/ (1-R _Tx)) * R _Tx); Wherein, R _TxBe transmission code rate, R _mBe female sign indicating number code check, m _b, n _bBe respectively the line number and the columns of unified basis matrix, z is a spreading factor, and func (x) is a kind of of number, round numbers downwards or the round numbers of rounding off that x is rounded up.

Said encoder matrix confirms that the unit specifically is used for said unified basis matrix is revised, and according to spreading factor z, utilizes the unit matrix of z * z that the basis matrix that correction obtains is expanded, and obtains the encoder matrix of LDPC sign indicating number.

Said information bit confirms that the comparative result that the unit specifically is used for when comparing unit is R _Tx>=R _mThe time, be that the information block bit of K ' is directly as information bit to be encoded with the length that receives; When the comparative result of comparing unit is R _Tx＜R _mThe time,, length adds filling bit in being the information block bit of K ', and formation length is k _bThe information bit to be encoded of * z.

Said code word processing unit specifically is used for female sign indicating number code word that the coding unit coding obtains is reset, and the female sign indicating number code word bits before wherein resetting is A ₀, A ₁..., A _{Nb * z-1}, the female sign indicating number code word bits after the rearrangement is B ₀, B ₁..., B _{Nb * z-1}Rearrangement formulae is:

Wherein, k _b=n _b-m _b, PV representes code word rearrangement vector, its element set is { n _b-m _b, n _b-m _b+ 1, n _b-m _b+ 2 ..., n _b-1}, index is the element of x in PV (x) the expression PV vector, x is a nonnegative integer;

When the comparative result of comparing unit is R _Tx>=R _mThe time, the female sign indicating number code word bits B after resetting ₀, B ₁..., B _{Nb * z-1}According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission;

When the comparative result of comparing unit is R _Tx＜R _mThe time, with the said k that adds _bThe female sign indicating number code word bits B of the individual filling bit of * z-K ' after reset ₀, B ₁..., B _{Nb * z-1}In systematic bits in delete, remaining female sign indicating number code word bits is as code word bits waiting for transmission.

Said encoder matrix confirms that the unit specifically is used for unified basis matrix is revised, and according to spreading factor z, utilizes the unit matrix of z * z that the basis matrix that correction obtains is expanded, and obtains the extended matrix of LDPC sign indicating number; When the comparative result of comparing unit is R _Tx>=R _mThe time, with the encoder matrix of extended matrix as the LDPC sign indicating number; When the comparative result of comparing unit is R _Tx＜R _mThe time, the func (m on intercepting extended matrix the right _b* z/ (1-R _Tx)) row, the matrix that intercepting is obtained is as the encoder matrix of LDPC sign indicating number; Or

During for positive integer, unified basis matrix is revised, obtained revised basis matrix; When the comparative result of comparing unit is R _Tx>=R _mThe time, according to spreading factor z, utilize the unit matrix of z * z that the basis matrix that correction obtains is expanded, obtain the encoder matrix of LDPC sign indicating number; When the comparative result of comparing unit is R _Tx＜R _mThe time, the k ' on the revised basis matrix of intercepting the right _b+ m _bRow according to spreading factor z, utilize the unit matrix of z * z that the matrix that intercepting obtains is expanded, and obtain the encoder matrix of LDPC sign indicating number.

Said information bit confirms that it is that the information block bit of K ' is directly as information bit to be encoded that the unit specifically is used for the length that receives.

Said code word processing unit specifically is used for female sign indicating number code word that coding unit obtains is reset, and the female sign indicating number code word bits after the female sign indicating number code word bits before wherein resetting is reset for

for

rearrangement formulae is:

wherein;

PV representes code word rearrangement vector; Its element set is the element of x for index in

PV (x) expression PV vector, and x is a nonnegative integer;

When the comparative result of comparing unit is R _Tx>=R _mThe time, the female sign indicating number code word bits after resetting

According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission;

When the comparative result of comparing unit is R _Tx＜R _mThe time, directly will be from the female sign indicating number code word bits after resetting

As code word bits waiting for transmission.

The coding method of LDPC sign indicating number provided by the invention and device; Confirm a basis matrix and spreading factor; Carry out the processing of confirming of input information block length or encoder matrix according to the transmission code rate that receives, be met the code word of aforementioned transmission code rate, on the basis of best performance, realized the changeability of code check; Reduced the complexity of hardware designs when realizing different code check, and realized that cost is lower.

Description of drawings

Fig. 1 is the structured flowchart of digital communication system;

Fig. 2 is the realization flow sketch map of the coding method of LDPC sign indicating number provided by the invention;

Fig. 3 is the schematic flow sheet of embodiment one of the coding method of LDPC sign indicating number provided by the invention;

Fig. 4 is the schematic flow sheet of embodiment two of the coding method of LDPC sign indicating number provided by the invention;

Fig. 5 is the structured flowchart of the code device of LDPC sign indicating number provided by the invention.

Embodiment

Basic thought of the present invention is: the female sign indicating number code check that relatively more unified basis matrix is supported and the size of transmission code rate, according to comparative result, confirm the length K of input information packet bit '; Confirm the encoder matrix of LDPC sign indicating number according to unifying basis matrix and spreading factor; Length by receiving is the information block bit of K ', confirms information bit to be encoded; Utilize said encoder matrix to treat the information encoded bit and carry out the LDPC coding, and female sign indicating number code word bits that coding obtains is handled, obtain code word bits waiting for transmission.

For making the object of the invention, technical scheme and advantage clearer, below lift embodiment and, the present invention is further explained with reference to accompanying drawing.

Fig. 2 shows the flow process of the coding method of LDPC sign indicating number provided by the invention, and is as shown in Figure 2, and said method comprises the steps:

S201, the unified basis matrix

and the spreading factor z of the LDPC sign indicating number of the fixed size that the reception user confirms;

In this step, the user can confirm according to the actual requirements that a size is m _b* n _bUnified basis matrix

With largest extension factor z _Max, m wherein _b, n _b, z _MaxBe positive integer, make k _b=n _b-m _b, then obtain

Female sign indicating number code check of supporting is R _m=(n _b-m _b)/n _b=k _b/ n _bAnd confirm spreading factor z according to the coding code length of digital communication system in transmission, usually, realize for the ease of hardware, spreading factor z be generally 2 power or 2 power positive integer doubly.

S202 receives current transmission code rate R _Tx, compare R _TxAnd R _mSize, according to comparative result, confirm the length K of input information packet bit ';

In this step, work as R _Tx>=R _mThe time, then the length K of information block '=k _b* z=(n _b-m _b) * z; Work as R _Tx＜R _mThe time, the m that then encodes and generate _bThe check bit of * z all can be transmitted, so information block length K ' can be by the check bit length m _b* z and transmission code rate R _TxConfirm, so K ', m _b* z and R _TxRelation can but be not limited to use following formulate:

Or

Or

Therefore, K ' can but be not limited to following formulate: K '=func (m _b* z * R _Tx/ (1-R _Tx)) or K '=func (m _b* z/ (1-R _Tx))-m _b* z or K '=func (func (m _b* z/ (1-R _Tx)) * R _Tx); Wherein func (x) can represent a kind of of number, round numbers downwards or the round numbers of rounding off that round up to x.

S203, according to spreading factor z with

Confirm the encoder matrix H of LDPC sign indicating number _Bz

In this step, at first utilize above-mentioned correction algorithm and z right

Revise, obtain revised basis matrix

According to

Obtain (m with the unit matrix expansion of z and z * z _b* z) * (n _b* z) LDPC matrix

Among the present invention both can with

Encoder matrix H as the LDPC sign indicating number _Bz, also can be according to R _TxAnd R _mComparative result, right

Carry out intercepting, obtain the encoder matrix H of LDPC sign indicating number _Bz

S204 is the information block bit of K ' by the length that receives, and confirms information bit to be encoded;

Particularly, according to encoder matrix H _BzSize and R _TxAnd R _mComparative result, the information block bit is handled, obtain information bit to be encoded.

S205 is through encoder matrix H _BzTreat the information encoded bit and carry out the LDPC coding, obtain female sign indicating number code word bits, female sign indicating number code word bits is handled, obtain code word bits waiting for transmission.

Particularly, according to transmission code rate R _TxCan confirm the length of code word bits waiting for transmission,, the surplus code word bits waiting for transmission that obtains is reset, shortens or deleted to female sign indicating number code word according to the length of the code word bits of confirming waiting for transmission.

Fig. 3 shows the realization flow of embodiment one of the coding method of LDPC sign indicating number provided by the invention, and as shown in Figure 3, said method comprises the steps:

S301～S302, the unified basis matrix of the LDPC sign indicating number of the fixed size that the reception user confirms

With largest extension factor z _Max, and confirm spreading factor z according to needed coding code length, carry out S305 then;

In this step, the user can confirm according to the actual requirements that a size is m _b* n _bUnified basis matrix

With largest extension factor z _Max, m wherein _b, n _b, z _MaxBe positive integer, make k _b=n _b-m _b, then obtain

Female sign indicating number code check of supporting is R _m=(n _b-m _b)/n _b=k _b/ n _bAnd confirm spreading factor z according to digital communication system needed coding code length in transmission, usually, realize for the ease of hardware, spreading factor z be generally 2 power or 2 power positive integer doubly.

S303～S304 receives current transmission code rate R _Tx, compare R _TxAnd R _mSize, according to comparative result, confirm the length K of input information packet bit ';

In this step, work as R _Tx>=R _mThe time, then the length K of information block '=k _b* z=(n _b-m _b) * z; Work as R _Tx＜R _mThe time, the m that then encodes and generate _bThe check bit of * z all can be transmitted, so information block length K ' can be by the check bit length m _b* z and transmission code rate R _TxConfirm, so K ', m _b* z and R _TxRelation can but be not limited to use following formulate:

Or

Or

Therefore, K ' can but be not limited to following formulate: K '=func (m _b* z * R _Tx/ (1-R _Tx)) or K '=func (m _b* z/ (1-R _Tx))-m _b* z or K '=func (func (m _b* z/ (1-R _Tx)) * R _Tx); Wherein func (x) can represent a kind of of number, round numbers downwards or the round numbers of rounding off that round up to x.

S305 is to unified basis matrix

Revise, utilize the z expansion to obtain the encoder matrix H of LDPC sign indicating number _Bz, carry out S308 then;

In this step, at first utilize above-mentioned correction algorithm and z right

Revise, obtain revised basis matrix

According to

Obtain (m with the unit matrix expansion of z and z * z _b* z) * (n _b* z) LDPC matrix

And with matrix Encoder matrix H as the LDPC sign indicating number _Bz

S306～S307 is the information block bit of K ' by the length that receives, and confirms information bit to be encoded;

Particularly, work as R _Tx>=R _mThe time, be that the information block bit of K ' is directly as information bit to be encoded with the length that receives; Work as R _Tx＜R _mThe time,, length adds filling bit in being the information block bit of K ', and formation length is k _bThe information bit to be encoded of * z, wherein add filling bit method can in information block bit front, back or mode such as alternate, the length of filling bit is k _b* z-K ', in addition, for subsequent treatment convenience, the k of filling _bThe individual bit of * z-K ' is preferably complete 0 bit, also can be complete 1 bit.

S308～S309 is through (m _b* z) * (n _b* z) encoder matrix H _BzTreat the information encoded bit and carry out the LDPC coding, obtain female sign indicating number code word bits, female sign indicating number code word bits is handled, obtain code word bits waiting for transmission.

Particularly, through encoder matrix H _BzLDPC coding, obtaining female sign indicating number code word bits length is N=n _b* z, the preceding k of wherein female sign indicating number code word bits _b* z bit is systematic bits, back m _b* z bit is check bit.

In order to improve the reliability of codeword transmission, can at first carry out code word and reset the female sign indicating number code word bits that obtains, the female sign indicating number code word bits before supposing to reset is A ₀, A ₁..., A _{Nb * z-1}, the female sign indicating number code word bits after code word is reset is B ₀, B ₁..., B _{Nb * z-1}, then code word is reset by following formulate:

Wherein, PV representes code word rearrangement vector, and PV comprises m _bIndividual different element, its element set are { n _b-m _b, n _b-m _b+ 1, n _b-m _b+ 2 ..., n _b-1}, and each element is different, index is the element of x in PV (x) the expression PV vector, wherein x is a nonnegative integer.

Work as R _Tx>=R _mThe time, delete surplus operation, promptly the female sign indicating number code word bits B that obtains after resetting ₀, B ₁..., B _{Nb * z-1}According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission, wherein func (x) can represent to x round up number, round numbers downwards or the round numbers of rounding off-kind, specifically can select suitable operation to func (x) according to the applicable cases of reality.

Work as R _Tx＜R _mThe time, shorten operation, with the k that adds among the S304 _bFemale sign indicating number code word bits B that the individual filling bit of * z-K ' obtains after reset ₀, B ₁,, B _{Nb * z-1}In systematic bits in delete (for the carrying out of deletion action here, thus among the S304 preferably filling bit be complete 0 or complete 1 bit), remaining female sign indicating number code word bits is as code word bits waiting for transmission.

Below in conjunction with example, the embodiment one of the coding method of LDPC sign indicating number is specified:

Example one:

1, receives the unified basis matrix of one 4 * 24 the LDPC sign indicating number that the user confirms With largest extension factor z _Max, z wherein _Max=96,

As follows:

$\begin{array}{cccccccccccccccccccccccc}1& 25& 55& -1& 47& 4& -1& 91& 84& 8& 86& 52& 82& 33& 5& 0& 36& 20& 4& 77& 80& 0& -1& -1\\ -1& 6& -1& 36& 40& 47& 12& 79& 47& -1& 41& 21& 12& 71& 14& 72& 0& 44& 49& 0& 0& 0& 0& -1\\ 51& 81& 83& 4& 67& -1& 21& -1& 31& 24& 91& 61& 81& 9& 86& 78& 60& 88& 67& 15& -1& -1& 0& 0\\ 68& -1& 50& 15& -1& 36& 13& 10& 11& 20& 53& 90& 29& 92& 57& 30& 84& 92& 11& 66& 80& -1& -1& 0\end{array}$

Can draw m _b=4, n _b=24, k _b=n _b-m _b=24-4=20,

Female sign indicating number code check of supporting is R _m=(n _b-m _b)/n _b=k _b/ n _b=20/24=5/6=0.833 for the ease of the coding code length of realization and actual demand, confirms spreading factor z=3 * 2 in addition ⁵=96.

2, receive current transmission code rate R _Tx=0.78, compare R _TxAnd R _mSize because 0.78＜0.833, therefore can pass through K '=ceil (m _b* z/ (1-R _Tx))-m _b* z, trying to achieve K ' is 1362, wherein ceil (x) expression is to the x number that rounds up.

3, through rounding (scale + floor) algorithm and extended correction factor z uniform basis matrix get corrected fundamental matrix

is as follows:

$\begin{array}{cccccccccccccccccccccccc}1& 25& 55& -1& 47& 4& -1& 91& 84& 8& 86& 52& 82& 33& 5& 0& 36& 20& 4& 77& 80& 0& -1& -1\\ -1& 6& -1& 36& 40& 47& 12& 79& 47& -1& 41& 21& 12& 71& 14& 72& 0& 44& 49& 0& 0& 0& 0& -1\\ 51& 81& 83& 4& 67& -1& 21& -1& 31& 24& 91& 61& 81& 9& 86& 78& 60& 88& 67& 15& -1& -1& 0& 0\\ 68& -1& 50& 15& -1& 36& 13& 10& 11& 20& 53& 90& 29& 92& 57& 30& 84& 92& 11& 66& 80& -1& -1& 0\end{array}$

According to

Expand the LDPC matrix that obtains (4 * 96) * (24 * 96) with the unit matrix of z=96 and z * z

And with matrix

Encoder matrix H as the LDPC sign indicating number _Bz

4, receive the information block bit that length is K ', because 0.78＜0.833, i.e. R _Tx＜R _m, be k with length _b* z-K '=558 complete 0 bit is filled into the front that length is the information block bit of K ', and obtaining length is k _bThe information bit to be encoded of * z=1920.

5, with the H of (4 * 96) * (24 * 96) _BzTo length is k _bThe information bit to be encoded of * z=1920 carries out the LDPC coding, and obtaining length is N=n _bFemale sign indicating number code word bits of * z=2304, wherein preceding 1920 bits are systematic bits, back 384 bits are check bit.

Female sign indicating number code word bits is carried out code word reset, the female sign indicating number code word bits before supposing to reset is A ₀, A ₁,, A ₂₃₀₃Female sign indicating number code word bits after code word is reset is B ₀, B ₁..., B ₂₃₀₃Then code word is reset by following formulate:

Wherein, PV representes code word rearrangement vector, PV={20,21,22,23}.Again because 0.78＜0.833, i.e. R _Tx＜R _m, female sign indicating number code word bits B that 558 complete 0 bits that are filled into information block bit front in 4 are obtained after reset ₀, B ₁..., B ₂₃₀₃In systematic bits in delete remaining female sign indicating number code word bits B ₅₅₈, B ₅₅₉..., B ₂₃₀₃As code word bits waiting for transmission.

Example two:

1, receives the unified basis matrix of one 4 * 24 the LDPC sign indicating number that the user confirms

With largest extension factor z _Max, z wherein _Max=96,

As follows:

$\begin{array}{cccccccccccccccccccccccc}1& 25& 55& -1& 47& 4& -1& 91& 84& 8& 86& 52& 82& 33& 5& 0& 36& 20& 4& 77& 80& 0& -1& -1\\ -1& 6& -1& 36& 40& 47& 12& 79& 47& -1& 41& 21& 12& 71& 14& 72& 0& 44& 49& 0& 0& 0& 0& -1\\ 51& 81& 83& 4& 67& -1& 21& -1& 31& 24& 91& 61& 81& 9& 86& 78& 60& 88& 67& 15& -1& -1& 0& 0\\ 68& -1& 50& 15& -1& 36& 13& 10& 11& 20& 53& 90& 29& 92& 57& 30& 84& 92& 11& 66& 80& -1& -1& 0\end{array}$

Can obtain m _b=4, n _b=24, k _b=n _b-m _b=24-4=20, Female sign indicating number code check of supporting is R _m=(n _b-m _b)/n _b=k _b/ n _b=20/24=5/6=0.833 for the ease of realizing, confirms spreading factor z=2 in addition ⁶=64.

2, receive current transmission code rate R _Tx=0.875, compare R _TxAnd R _mSize because 0.875＞0.833, so information block length K '=k _b* z=1280.

3, through rounding (scale + floor) algorithm and extended correction factor z uniform basis matrix

get corrected fundamental matrix

is as follows:

$\begin{array}{cccccccccccccccccccccccc}0& 16& 36& -1& 31& 2& -1& 60& 56& 5& 57& 34& 54& 22& 3& 0& 24& 13& 2& 51& 53& 0& -1& -1\\ -1& 4& -1& 34& 26& 31& 8& 52& 31& -1& 27& 14& 8& 47& 9& 48& 0& 29& 32& 0& 0& 0& 0& -1\\ 34& 54& 55& 2& 44& -1& 14& -1& 20& 16& 60& 40& 54& 6& 57& 52& 40& 58& 44& 10& -1& -1& 0& 0\\ 45& -1& 33& 10& -1& 24& 8& 6& 7& 13& 35& 60& 19& 61& 38& 20& 56& 61& 7& 44& 53& -1& -1& 0\end{array}$

According to Expand the LDPC matrix that obtains (4 * 64) * (24 * 64) with the unit matrix of z=64 and z * z

And with matrix

Encoder matrix H as the LDPC sign indicating number _Bz

4, receive the information block bit that length is K '=1280, because 0.875＞0.833, be K '=k with length _b* z=1280 information block bit is directly as information bit to be encoded.

5, with the H of (4 * 64) * (24 * 64) _BzTo length is K '=k _bThe information bit to be encoded of * z=1280 carries out the LDPC coding, and obtaining length is N=n _bFemale sign indicating number code word bits of * z=1536, wherein preceding 1280 bits are systematic bits, back 256 bits are check bit.

Female sign indicating number code word bits is carried out code word reset, the female sign indicating number code word bits before supposing to reset is A ₀, A ₁..., A ₁₅₃₅, the female sign indicating number code word bits after code word is reset is B ₀, B ₁..., B ₁₅₃₅, then code word is reset by following formulate:

Wherein, PV representes code word rearrangement vector, PV={20,21,22,23}.Owing to 0.875＞0.833, belong to R again _Tx>=R _m, female sign indicating number code word bits B that rearrangement is obtained ₀, B ₁..., B ₁₅₃₅According to index get successively from small to large floor (K '/R _Tx)=1462 bit is promptly B ₀, B ₁..., B ₁₄₆₁As code word bits waiting for transmission, wherein, floor (x) representes the downward round numbers of x.

Fig. 4 shows the realization flow of embodiment two of the coding method of LDPC sign indicating number provided by the invention, and as shown in Figure 4, said method comprises the steps:

S401～S402, the unified basis matrix of the LDPC sign indicating number of the fixed size that the reception user confirms

With largest extension factor z _Max, and confirm spreading factor z according to needed coding code length;

In this step, the user can confirm according to the actual requirements that a size is m _b* n _bUnified basis matrix With largest extension factor z _Max, m wherein _b, n _b, z _MaxBe positive integer, make k _b=n _b-m _b, then obtain

Female sign indicating number code check of supporting is R _m=(n _b-m _b)/n _b=k _b/ n _bAnd confirm spreading factor z according to the coding code length of digital communication system in transmission, usually, realize for the ease of hardware, spreading factor z be generally 2 power or 2 power positive integer doubly.

S403 receives current transmission code rate R _Tx, carry out S404 and S405 then respectively.

S404, relatively R _TxAnd R _mSize, according to comparative result, confirm the length K of input information packet bit ', carry out S406 then;

In this step, work as R _Tx>=R _mThe time, then the length K of information block '=k _b* z=(n _b-m _b) * z; Work as R _Tx＜R _mThe time, the m that then encodes and generate _bThe check bit of * z all can be transmitted, so information block length K ' can be by the check bit length m _b* z and transmission code rate R _TxConfirm, so K ', m _b* z and R _TxRelation can but be not limited to use following formulate:

Or

Or

Therefore, K ' can but be not limited to following formulate: K '=func (m _b* z * R _Tx/ (1-R _Tx)) or K '=func (m _b* z/ (1-R _Tx))-m _b* z or K '=func (func (m _b* z/ (1-R _Tx)) * R _Tx); Wherein func (x) can represent a kind of of number, round numbers downwards or the round numbers of rounding off that round up to x.

S405 is to unified basis matrix

Revise, confirm the encoder matrix H of LDPC sign indicating number _Bz, carry out S407 then;

In this step, at first utilize above-mentioned correction algorithm and z right

Revise, obtain revised basis matrix

According to

Obtain (m with the unit matrix expansion of z and z * z _b* z) * (n _b* z) LDPC extended matrix

Work as R _Tx>=R _mThe time, then with extended matrix

Encoder matrix H as the LDPC sign indicating number _BzWork as R _Tx＜R _mThe time, through intercepting

Func (the m on the right _b* z/ (1-R _Tx)) row, obtain size and be (m _b* z) * func (m _b* z/ (1-R _Tx)) encoder matrix H _BzAgain

Therefore, this moment the LDPC sign indicating number encoder matrix H _BzSize be (m _b* z) * (K '+m _b* z).

In addition, when

During for positive integer, in this step to unified basis matrix

Revise and obtain After, also can be according to R _TxAnd R _mComparative result, to revised basis matrix

After handling, utilize spreading factor z to expand, obtain the encoder matrix H of LDPC sign indicating number _Bz

Particularly, work as R _Tx>=R _mThe time, to revised basis matrix Expand, obtain encoder matrix H _Bz, concrete and said method obtains encoder matrix H _BzProcess identical, repeat no more;

Work as R _Tx＜R _mThe time, through intercepting The k ' on the right _b+ m _bRow obtain size and are m _b* (k ' _b+ m _b) matrix

According to

With spreading factor z, expansion obtains (m _b* z) * ((k ' _b+ m _bThe encoder matrix H of the LDPC sign indicating number of) * z) _BzAmong the S403, work as R again _Tx＜R _mThe time,

Again Be positive integer, therefore can obtain,

Promptly

So the time, the encoder matrix H of the LDPC sign indicating number that obtains _BzSize also be (m _b* z) * (K '+m _b* z).

S406, receiving length is the information block bit of K ', and with the information block bit directly as information bit to be encoded;

S407～S408 is through encoder matrix H _BzTreat the information encoded bit and carry out the LDPC coding, obtain female sign indicating number code word bits, female sign indicating number code word bits is handled, obtain code word bits waiting for transmission.

Particularly, work as R _Tx>=R _mThe time, be (m through size _b* z) * (n _b* z) the encoder matrix matrix H of LDPC sign indicating number _BzTreat the information encoded bit and carry out the LDPC coding, the length of the female sign indicating number code word bits that obtains is N=K '+m _b* z=n _b* z, the individual bit of preceding K ' of wherein female sign indicating number code word bits is a systematic bits, back m _b* z bit is check bit, and wherein, information bit length to be encoded is K '=k _b* z=(n _b-m _b) * z.

In order to improve the reliability of codeword transmission; At first a female sign indicating number code word bits that obtains is carried out code word and resets, the female sign indicating number code word bits before suppose to reset for the female sign indicating number code word bits of

code word after resetting for then the code word rearrangement by following formulate:

Wherein,

PV representes code word rearrangement vector, and PV comprises m _bIndividual different element, its element set does

And each element is different, and index is the element of x in PV (x) the expression PV vector, and wherein x is a nonnegative integer.

Work as R _Tx>=R _mThe time, delete surplus operation, promptly the female sign indicating number code word bits that obtains after resetting

According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission, promptly Be code word bits waiting for transmission.Wherein func (x) can represent a kind of of number, round numbers downwards or the round numbers of rounding off that round up to x, specifically can select suitable operation to func (x) according to the applicable cases of reality.

Work as R _Tx＜R _mThe time, owing to obtaining encoder matrix H among the S403 _BzCarried out intercept operation in the process, so can be with the female sign indicating number code word bits that obtains after resetting

As code word bits waiting for transmission.

Below in conjunction with example, the embodiment two of the coding method of LDPC sign indicating number is specified:

Example three:

1, receives the unified basis matrix of one 6 * 24 the LDPC sign indicating number that the user confirms With largest extension factor z _Max, z wherein _Max=96,

As follows:

$\begin{array}{cccccccccccccccccccccccc}6& 38& 3& 93& -1& -1& -1& 30& 70& -1& 86& -1& 37& 30& 4& 11& -1& 46& 48& 0& -1& -1& -1& -1\\ 62& 94& 19& 84& -1& 92& 78& -1& 15& -1& -1& 92& -1& 45& 24& 32& 30& -1& -1& 0& 0& -1& -1& -1\\ 71& -1& 55& -1& 12& 66& 45& 79& -1& 78& -1& -1& 10& -1& 22& 55& 70& 82& -1& -1& 0& 0& -1& -1\\ 38& 61& -1& 66& 9& 73& 47& 64& -1& 39& 61& 43& -1& -1& -1& -1& 95& 32& 0& -1& -1& 0& 0& -1\\ -1& -1& -1& -1& 32& 52& 55& 80& 95& 22& 6& 51& 24& 90& 44& 20& -1& -1& -1& -1& -1& -1& 0& 0\\ -1& 63& 31& 88& 20& -1& -1& -1& 6& 40& 56& 16& 71& 53& -1& -1& 27& 26& 48& -1& -1& -1& -1& 0\end{array}$

Can draw m _b=6, n _b=24, k _b=n _b-m _b=24-6=18,

Female sign indicating number code check of supporting is R _m=(n _b-m _b)/n _b=k _b/ n _b=18/24=3/4=0.75 for the ease of the coding code length of realization and actual demand, confirms spreading factor z=3 * 2 in addition ⁵=96.

2, receive current transmission code rate R _Tx=0.9, compare R _TxAnd R _mSize because 0.9＞0.75, so information block length K '=k _b* z=1728.

3, through rounding (scale + floor) algorithm and extended correction factor z uniform basis matrix

get corrected fundamental matrix

is as follows:

$\begin{array}{cccccccccccccccccccccccc}6& 38& 3& 93& -1& -1& -1& 30& 70& -1& 86& -1& 37& 30& 4& 11& -1& 46& 48& 0& -1& -1& -1& -1\\ 62& 94& 19& 84& -1& 92& 78& -1& 15& -1& -1& 92& -1& 45& 24& 32& 30& -1& -1& 0& 0& -1& -1& -1\\ 71& -1& 55& -1& 12& 66& 45& 79& -1& 78& -1& -1& 10& -1& 22& 55& 70& 82& -1& -1& 0& 0& -1& -1\\ 38& 61& -1& 66& 9& 73& 47& 64& -1& 39& 61& 43& -1& -1& -1& -1& 95& 32& 0& -1& -1& 0& 0& -1\\ -1& -1& -1& -1& 32& 52& 55& 80& 95& 22& 6& 51& 24& 90& 44& 20& -1& -1& -1& -1& -1& -1& 0& 0\\ -1& 63& 31& 88& 20& -1& -1& -1& 6& 40& 56& 16& 71& 53& -1& -1& 27& 26& 48& -1& -1& -1& -1& 0\end{array}$

According to Expand the LDPC matrix that obtains (6 * 96) * (24 * 96) with the unit matrix of z=96 and z * z

Again 0.9＞0.75, then with matrix Encoder matrix H as the LDPC sign indicating number _Bz

4, receive the information block bit that length is K '=1728, and be K '=k length _b* z=1728 information block bit is directly as information bit to be encoded.

5, with the H of (6 * 96) * (24 * 96) _BzTo length is K '=k _bThe information bit to be encoded of * z=1728 carries out the LDPC coding, and obtaining length is N=K '+m _b* z=n _bFemale sign indicating number code word bits of * z=2304, wherein preceding 1728 bits are systematic bits, back 576 bits are check bit.

Female sign indicating number code word bits is carried out code word reset, the female sign indicating number code word bits before supposing to reset is A ₀, A ₁..., A ₂₃₀₃, the female sign indicating number code word bits after code word is reset is B ₀, B ₁..., B ₂₃₀₃, then code word is reset by following formulate:

Wherein, PV representes code word rearrangement vector, PV={18,19; 20; 21,22,23}.Owing to 0.9＞0.75, belong to R again _Tx>=R _m, female sign indicating number code word bits B that rearrangement is obtained ₀, B ₁..., B ₂₃₀₃According to index get successively from small to large ceil (K '/R _Tx)=1920 bit is promptly B ₀, B ₁..., B ₁₉₁₉As code word bits waiting for transmission, wherein, ceil (x) expression is to the x number that rounds up.

Example four:

1, receives the unified basis matrix of one 6 * 24 the LDPC sign indicating number that the user confirms

With largest extension factor z _Max, z wherein _Max=96,

As follows:

$\begin{array}{cccccccccccccccccccccccc}6& 38& 3& 93& -1& -1& -1& 30& 70& -1& 86& -1& 37& 30& 4& 11& -1& 46& 48& 0& -1& -1& -1& -1\\ 62& 94& 19& 84& -1& 92& 78& -1& 15& -1& -1& 92& -1& 45& 24& 32& 30& -1& -1& 0& 0& -1& -1& -1\\ 71& -1& 55& -1& 12& 66& 45& 79& -1& 78& -1& -1& 10& -1& 22& 55& 70& 82& -1& -1& 0& 0& -1& -1\\ 38& 61& -1& 66& 9& 73& 47& 64& -1& 39& 61& 43& -1& -1& -1& -1& 95& 32& 0& -1& -1& 0& 0& -1\\ -1& -1& -1& -1& 32& 52& 55& 80& 95& 22& 6& 51& 24& 90& 44& 20& -1& -1& -1& -1& -1& -1& 0& 0\\ -1& 63& 31& 88& 20& -1& -1& -1& 6& 40& 56& 16& 71& 53& -1& -1& 27& 26& 48& -1& -1& -1& -1& 0\end{array}$

Can draw m _b=6, n _b=24, k _b=n _b-m _b=24-6=18,

Female sign indicating number code check of supporting is R _m=(n _b-m _b)/n _b=k _b/ n _b=18/24=3/4=0.75 for the ease of the coding code length of realization and actual demand, confirms spreading factor z=2 in addition ⁶=64.

2, receive current transmission code rate R _Tx=0.6, compare R _TxAnd R _mSize because 0.6＜0.75, therefore can pass through K '=ceil (m _b* z/ (1-R _Tx))-m _b* z, trying to achieve K ' is 576, wherein ceil (x) expression is to the x number that rounds up.

3, through rounding (scale + floor) algorithm and extended correction factor z uniform basis matrix

get corrected fundamental matrix

is as follows:

$\begin{array}{cccccccccccccccccccccccc}4& 25& 2& 62& -1& -1& -1& 20& 46& -1& 57& -1& 24& 25& 2& 7& -1& 30& 32& 0& -1& -1& -1& -1\\ 41& 62& 12& 56& -1& 61& 52& -1& 10& -1& -1& 61& -1& 30& 16& 21& 20& -1& -1& 0& 0& -1& -1& -1\\ 47& -1& 36& -1& 8& 44& 30& 52& -1& 52& -1& -1& 6& -1& 14& 36& 46& 54& -1& -1& 0& 0& -1& -1\\ 25& 40& -1& 44& 6& 48& 31& 42& -1& 26& 40& 28& -1& -1& -1& -1& 63& 21& 0& -1& -1& 0& 0& -1\\ -1& -1& -1& -1& 21& 34& 36& 53& 63& 14& 4& 34& 16& 60& 29& 13& -1& -1& -1& -1& -1& -1& 0& 0\\ -1& 42& 20& 58& 13& -1& -1& -1& 4& 26& 37& 10& 47& 35& -1& -1& 18& 17& 32& -1& -1& -1& -1& 0\end{array}$

According to

Expand the LDPC matrix that obtains (6 * 64) * (24 * 64) with the unit matrix of z=64 and z * z Again because 0.6＜0.75, so intercepting

Ceil (the m on the right _b* z/ (1-R _Tx))=960 row, obtain size and be the encoder matrix H of (6 * 64) * 960 _Bz

In addition, this step can also be handled like this: obtain revised basis matrix

Again because 0.6＜0.75,

Intercepting

The k ' on the right _b+ m _b=9+6=15 row obtain size and are 6 * (9+6) matrix

According to

Expand the LDPC encoder matrix H that obtains (6 * 64) * (15 * 64) with the unit matrix of z=64 and z * z _Bz

4, receive the information block bit that length is K ', and be that the information block bit of K '=576 is directly as information bit to be encoded length.

5, with the H of (6 * 64) * (15 * 64) _BzTo length is that the information bit to be encoded of K '=576 carries out the LDPC coding, and obtaining length is N=K '+m _b* z=n _bFemale sign indicating number code word bits of * z=960, wherein preceding 576 bits are systematic bits, back 384 bits are check bit.

Female sign indicating number code word bits is carried out code word reset, the female sign indicating number code word bits before supposing to reset is A ₀, A ₁..., A ₉₅₉, the female sign indicating number code word bits after code word is reset is B ₀, B ₁..., B ₉₅₉, then code word is reset by following formulate:

Wherein, PV representes code word rearrangement vector, PV={9,10; 11; 12,13,14}.Again because 0.6＜0.75, i.e. R _Tx＜R _m, owing to obtaining encoder matrix H in 3 _BzCarried out intercept operation in the process, so can be with the female sign indicating number code word bits B that obtains after resetting ₀, B ₁..., B ₉₅₉As code word bits waiting for transmission.

Fig. 5 shows the structure of the code device of LDPC sign indicating number provided by the invention; As shown in Figure 5, said device comprises: coding unit 10, comparing unit 20, packet bit confirm that unit 30, encoder matrix confirm that unit 40, information bit confirm unit 50 and code word processing unit 60; Wherein, comparing unit 20 is used for the female sign indicating number code check of relatively more unified basis matrix support and the size of transmission code rate; Packet bit confirms that unit 30 is used for the comparative result according to comparing unit, confirms the length K of input information packet bit '; Encoder matrix confirms that unit 40 is used for confirming the encoder matrix of LDPC sign indicating number according to unifying basis matrix and spreading factor; Information bit confirms that unit 50 is used to receive the information bit of input; Confirm the information block bit length that unit 30 is confirmed according to packet bit; Obtaining length is the information block bit of K ', and confirms information bit to be encoded, and it is transferred to coding unit 10; Code word processing unit 60 is used for that coding unit 10 is carried out the female sign indicating number code word that LDPC coding obtains to be handled, and obtains code word bits waiting for transmission.

Further, said device also comprises unit 70 is set, and is used to be provided with the unified basis matrix and the spreading factor of LDPC sign indicating number.

Further, said device also comprises input unit 80, is used to receive current transmission code rate; Packet bit confirms that the comparative result that unit 30 specifically is used for when comparing unit 20 is R _Tx>=R _mThe time, confirm the length K of information block bit ' be (n _b-m _b) * z; When the comparative result of comparing unit 20 is R _Tx＜R _mThe time, the length of confirming the information block bit is K '=func (m _b* z * R _Tx/ (1-R _Tx)) or K '=func (m _b* z/ (1-R _Tx))-m _b* z or K '=func (func (m _b* z/ (1-R _Tx)) * R _Tx); Wherein, R _TxBe transmission code rate, R _mBe female sign indicating number code check, m _b, n _bBe respectively the line number and the columns of unified basis matrix, z is a spreading factor, and func (x) is a kind of of number, round numbers downwards or the round numbers of rounding off that x is rounded up.

Further, encoder matrix confirms that unit 40 specifically is used for said unified basis matrix is revised, and according to spreading factor z, utilizes the unit matrix of z * z that the basis matrix that correction obtains is expanded, and obtains the encoder matrix of LDPC sign indicating number.

Correspondingly, information bit confirms that the comparative result that unit 50 specifically is used for when comparing unit 20 is R _Tx>=R _mThe time, be that the information block bit of K ' is directly as information bit to be encoded with the length that receives; When the comparative result of comparing unit 20 is R _Tx＜R _mThe time,, length adds filling bit in being the information block bit of K ', and formation length is k _bThe information bit to be encoded of * z, wherein filling bit is complete 0 bit or complete 1 bit.

Correspondingly, code word processing unit 60 specifically is used for female sign indicating number code word that coding unit 10 codings obtain is reset, and the female sign indicating number code word bits before wherein resetting is A ₀, A ₁..., A _{Nb * z-1}Female sign indicating number code word bits after the rearrangement is B ₀, B ₁..., B _{Nb * z-1}, rearrangement formulae is:

Wherein, k _b=n _b-m _b, PV representes code word rearrangement vector, its element set is { n _b-m _b, n _b-m _b+ 1, n _b-m _b+ 2 ..., n _b-1}, index is the element of x in PV (x) the expression PV vector, x is a nonnegative integer; When the comparative result of comparing unit 20 is R _Tx>=R _mThe time, the female sign indicating number code word bits B after resetting ₀, B ₁..., B _{Nb * z-1}According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission; When the comparative result of comparing unit 20 is R _Tx＜R _mThe time, with the said k that adds _bThe female sign indicating number code word bits B of the individual filling bit of * z-K ' after reset ₀, B ₁..., B _{Nb * z-1}In systematic bits in delete, remaining female sign indicating number code word bits is as code word bits waiting for transmission.

Further, encoder matrix confirms that unit 40 specifically is used for unified basis matrix is revised, and according to spreading factor z, utilizes the unit matrix of z * z that the basis matrix that correction obtains is expanded, and obtains the extended matrix of LDPC sign indicating number; When the comparative result of comparing unit 20 is R _Tx>=R _mThe time, with the encoder matrix of extended matrix as the LDPC sign indicating number; When the comparative result of comparing unit is R _Tx＜R _mThe time, the func (m on intercepting extended matrix the right _b* z/ (1-R _Tx)) row, the matrix that intercepting is obtained is as the encoder matrix of LDPC sign indicating number; Or During for positive integer, unified basis matrix is revised, obtained revised basis matrix; When the comparative result of comparing unit 20 is R _tX>=R _mThe time, according to spreading factor z, utilize the unit matrix of z * z that the basis matrix that correction obtains is expanded, obtain the encoder matrix of LDPC sign indicating number; When the comparative result of comparing unit 20 is R _Tx＜R _mThe time, the k ' on the revised basis matrix of intercepting the right _b+ m _bRow according to spreading factor z, utilize the unit matrix of z * z that the matrix that intercepting obtains is expanded, and obtain the encoder matrix of LDPC sign indicating number.

Correspondingly, information bit confirms that it is that the information block bit of K ' is directly as information bit to be encoded that unit 50 specifically is used for the length that receives.

Correspondingly, code word processing unit 60 specifically is used for female sign indicating number code word that coding unit 10 obtains is reset, and the female sign indicating number code word bits before wherein resetting does

Female sign indicating number code word bits after the rearrangement does

Rearrangement formulae is:

Wherein,

PV representes code word rearrangement vector, and its element set does

Index is the element of x in PV (x) the expression PV vector, and x is a nonnegative integer; When the comparative result of comparing unit 20 is R _Tx>=R _mThe time, the female sign indicating number code word bits after resetting

According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission; When the comparative result of comparing unit 20 is R _Tx＜R _mThe time, directly will be from the female sign indicating number code word bits after resetting

As code word bits waiting for transmission.

Other needs explanation, and encoding and decoding is corresponding each other, so the coding method and the device of LDPC sign indicating number provided by the invention, and its coding/decoding method and decoding device also have corresponding inverse operation, repeat no more.

The above is merely preferred embodiment of the present invention, is not to be used to limit protection scope of the present invention.

Claims (18)

1. the coding method of a low-density checksum LDPC sign indicating number is characterized in that, said method comprises the steps:

The female sign indicating number code check that relatively more unified basis matrix is supported and the size of transmission code rate based on comparative result, are confirmed the length K of input information block bit ';

Confirm the encoder matrix of LDPC sign indicating number according to unifying basis matrix and spreading factor;

Length by receiving is the information block bit of K ', confirms information bit to be encoded;

Utilize said encoder matrix to treat the information encoded bit and carry out the LDPC coding, and female sign indicating number code word bits that coding obtains is handled, obtain code word bits waiting for transmission.

2. method according to claim 1 is characterized in that, said method also comprises: unified basis matrix and spreading factor that the LDPC sign indicating number is set.

3. based on claim 1 or 2 described methods, it is characterized in that, also comprise before the female sign indicating number code check of said relatively more unified basis matrix support and the size of transmission code rate: receive current transmission code rate;

Said according to comparative result, confirm the length K of input information packet bit ' be:

Work as R _Tx>=R _mThe time, confirm the length K of information block ' be (n _b-m _b) * z; Work as R _Tx＜R _mThe time, the length of confirming the information block bit is K '=func (m _b* z * R _Tx/ (1-R _Tx)) or K '=func (m _b* z/ (1-R _Tx))-m _b* z or K '=func (func (m _b* z/ (1-R _Tx)) * R _Tx);

Wherein, R _TxBe transmission code rate, R _mBe female sign indicating number code check, m _b, n _bBe respectively the line number and the columns of unified basis matrix, z is a spreading factor, and func (x) is a kind of of number, round numbers downwards or the round numbers of rounding off that x is rounded up.

4. method according to claim 1 and 2 is characterized in that, and is said according to unified basis matrix and spreading factor, confirms that the encoder matrix of LDPC sign indicating number is:

Unified basis matrix is revised,, utilized the unit matrix of z * z that the basis matrix that correction obtains is expanded, obtain the encoder matrix of LDPC sign indicating number according to spreading factor z.

5. method according to claim 4 is characterized in that, said is the information block bit of K ' by the length that receives, and confirms that information bit to be encoded is:

Work as R _Tx>=R _mThe time, be that the information block bit of K ' is directly as information bit to be encoded with the length that receives; Work as R _Tx＜R _mThe time,, length adds filling bit in being the information block bit of K ', and formation length is k _bThe information bit to be encoded of * z.

6. method according to claim 5 is characterized in that, said female sign indicating number code word bits that coding is obtained is treated to:

Female sign indicating number code word to coding obtains is reset, and the female sign indicating number code word bits before wherein resetting is A ₀, A ₁..., A _{Nb * z-1}Female sign indicating number code word bits after the rearrangement is B ₀, B ₁..., B _{Nb * z-1}Rearrangement formulae is:

Wherein, k _b=n _b-m _b, PV representes code word rearrangement vector, its element set is { n _b-m _b, n _b-m _b+ 1, n _b-m _b+ 2 ..., n _b-1}, index is the element of x in PV (x) the expression PV vector, x is a nonnegative integer;

Work as R _Tx>=R _mThe time, the female sign indicating number code word bits B after resetting ₀, B ₁..., B _{Nb * z-1}According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission;

Work as R _Tx＜R _mThe time, with the said k that adds _bThe female sign indicating number code word bits B of the individual filling bit of * z-K ' after reset ₀, B ₁..., B _{Nb * z-1}In systematic bits in delete, remaining female sign indicating number code word bits is as code word bits waiting for transmission.

7. method according to claim 3 is characterized in that, the encoder matrix of said definite LDPC sign indicating number is:

Unified basis matrix is revised,, utilized the unit matrix of z * z that the basis matrix that correction obtains is expanded, obtain the extended matrix of LDPC sign indicating number according to spreading factor z;

Work as R _Tx>=R _mThe time, with the encoder matrix of extended matrix as the LDPC sign indicating number; Work as R _Tx＜R _mThe time, the func (m on intercepting extended matrix the right _b* z/ (1-R _Tx)) row, the matrix that intercepting is obtained is as the encoder matrix of LDPC sign indicating number; Perhaps,

As

when being positive integer; Unified basis matrix is revised, obtained revised basis matrix;

Work as R _Tx>=R _mThe time, according to spreading factor z, utilize the unit matrix of z * z that the basis matrix that correction obtains is expanded, obtain the encoder matrix of LDPC sign indicating number; Work as R _Tx＜R _mThe time, the k ' on the revised basis matrix of intercepting the right _b+ m _bRow according to spreading factor z, utilize the unit matrix of z * z that the matrix that intercepting obtains is expanded, and obtain the encoder matrix of LDPC sign indicating number.

8. method according to claim 7 is characterized in that, said is the information block bit of K ' by the length that receives, and confirms that information bit to be encoded is: with length is that the information block bit of K ' is directly as information bit to be encoded.

9. method according to claim 8 is characterized in that, said female sign indicating number code word bits that coding is obtained is treated to:

Female sign indicating number code word to coding obtains is reset, and the female sign indicating number code word bits after the female sign indicating number code word bits before wherein resetting is reset for for

rearrangement formulae is:

wherein;

PV representes code word rearrangement vector; Its element set is the element of x for index in PV (x) expression PV vector; X is a nonnegative integer, and wherein

expression is to the x number that rounds up;

Work as R _Tx>=R _mThe time, the female sign indicating number code word bits after resetting

According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission; Work as R _Tx＜R _mThe time, directly will be from the female sign indicating number code word bits after resetting

As code word bits waiting for transmission.

10. the code device of a LDPC sign indicating number comprises coding unit, it is characterized in that, said device also comprises: comparing unit, packet bit confirm that unit, encoder matrix confirm that unit, information bit confirm unit and code word processing unit; Wherein,

Comparing unit is used for the female sign indicating number code check of relatively more unified basis matrix support and the size of transmission code rate;

Packet bit is confirmed the unit, is used for the comparative result according to comparing unit, confirms the length K of input information packet bit ';

Encoder matrix is confirmed the unit, is used for confirming the encoder matrix of LDPC sign indicating number based on unifying basis matrix and spreading factor;

Information bit is confirmed the unit, is used for receiving obtaining the information block bit that length is K ', and confirms information bit to be encoded, and it is transferred to coding unit;

The code word processing unit is used for that coding unit is carried out the female sign indicating number code word that LDPC coding obtains and handles, and obtains code word bits waiting for transmission.

11. device according to claim 10 is characterized in that, said device also comprises: the unit is set, is used to be provided with the unified basis matrix and the spreading factor of LDPC sign indicating number.

12. according to claim 10 or 11 described devices, it is characterized in that said device also comprises input unit, be used to receive current transmission code rate;

Said packet bit confirms that the comparative result that the unit specifically is used for when comparing unit is R _Tx>=R _mThe time, confirm the length K of information block bit ' be (n _b-m _b) * z; When the comparative result of comparing unit is R _Tx＜R _mThe time, the length of confirming the information block bit is K '=func (m _b* z * R _Tx/ (1-R _Tx)) or K '=func (m _b* z/ (1-R _Tx))-m _b* z or K '=func (func (m _b* z/ (1-R _Tx)) * R _Tx); Wherein, R _TxBe transmission code rate, R _mBe female sign indicating number code check, m _b, n _bBe respectively the line number and the columns of unified basis matrix, z is a spreading factor, and func (x) is a kind of of number, round numbers downwards or the round numbers of rounding off that x is rounded up.

13. according to claim 10 or 11 described devices; It is characterized in that; Said encoder matrix confirms that the unit specifically is used for said unified basis matrix is revised; According to spreading factor z, utilize the unit matrix of z * z that the basis matrix that correction obtains is expanded, obtain the encoder matrix of LDPC sign indicating number.

14. device according to claim 13 is characterized in that, said information bit confirms that the comparative result that the unit specifically is used for when comparing unit is R _Tx>=R _mThe time, be that the information block bit of K ' is directly as information bit to be encoded with the length that receives; When the comparative result of comparing unit is R _Tx＜R _mThe time,, length adds filling bit in being the information block bit of K ', and formation length is k _bThe information bit to be encoded of * z.

15. device according to claim 14 is characterized in that, said code word processing unit specifically is used for female sign indicating number code word that the coding unit coding obtains is reset, and the female sign indicating number code word bits before wherein resetting is A ₀, A ₁..., A _{Nb * z-1}, the female sign indicating number code word bits after the rearrangement is B ₀, B ₁..., n _{B * z-1}Rearrangement formulae is:

Wherein, k _b=n _b-m _b, PV representes code word rearrangement vector, its element set is { n _b-m _b, n _b-m _b+ 1, n _b-m _b+ 2 ..., n _b-1}, index is the element of x in PV (x) the expression PV vector, x is a nonnegative integer;

When the comparative result of comparing unit is R _Tx>=R _mThe time, the female sign indicating number code word bits B after resetting ₀, B ₁..., B _{Nb * z-1}According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission;

When the comparative result of comparing unit is R _Tx＜R _mThe time, with the said k that adds _bThe female sign indicating number code word bits B of the individual filling bit of * z-K ' after reset ₀, B ₁..., B _{Nb * z-1}In systematic bits in delete, remaining female sign indicating number code word bits is as code word bits waiting for transmission.

16. device according to claim 12; It is characterized in that said encoder matrix confirms that the unit specifically is used for unified basis matrix is revised, according to spreading factor z; Utilize the unit matrix of z * z that the basis matrix that correction obtains is expanded, obtain the extended matrix of LDPC sign indicating number; When the comparative result of comparing unit is R _Tx>=R _mThe time, with the encoder matrix of extended matrix as the LDPC sign indicating number; When the comparative result of comparing unit is R _Tx＜R _mThe time, the func (m on intercepting extended matrix the right _b* z/ (1-R _Tx)) row, the matrix that intercepting is obtained is as the encoder matrix of LDPC sign indicating number; Or During for positive integer, unified basis matrix is revised, obtained revised basis matrix; When the comparative result of comparing unit is R _Tx>=R _mThe time, according to spreading factor z, utilize the unit matrix of z * z that the basis matrix that correction obtains is expanded, obtain the encoder matrix of LDPC sign indicating number; When the comparative result of comparing unit is R _Tx＜R _mThe time, the k ' on the revised basis matrix of intercepting the right _b+ m _bRow according to spreading factor z, utilize the unit matrix of z * z that the matrix that intercepting obtains is expanded, and obtain the encoder matrix of LDPC sign indicating number.

17. device according to claim 16 is characterized in that, said information bit confirms that it is that the information block bit of K ' is directly as information bit to be encoded that the unit specifically is used for the length that receives.

18. device according to claim 17; It is characterized in that; Said code word processing unit specifically is used for female sign indicating number code word that coding unit obtains is reset, and the female sign indicating number code word bits after the female sign indicating number code word bits before wherein resetting is reset for for

rearrangement formulae is:

wherein;

PV representes code word rearrangement vector; Its element set is the element of x for index in PV (x) expression PV vector, and x is a nonnegative integer;

When the comparative result of comparing unit is R _Tx>=R _mThe time, the female sign indicating number code word bits after resetting

According to index get successively from small to large func (K '/R _Tx) individual bit, as code word bits waiting for transmission;

When the comparative result of comparing unit is R _Tx＜R _mThe time, directly will be from the female sign indicating number code word bits after resetting

As code word bits waiting for transmission.

Images