CN1252935C - Information source-channel united coding method based on low-density odd-even check coding - Google Patents

Abstract

The source-channel joint coding method based on low-density parity-check coding belongs to the field of communication technology, and is characterized in that it is a source-channel joint coding method that combines hidden Markov source estimation and low-density parity-check (LDPC) coding The coding method, at the sending end, performs LDPC encoding on the source output sequence containing redundancy and then modulates the output, and at the receiving end, adopts the decoding method of joint source estimation and channel decoding, that is: first calculate the hard decision of the receiving sequence, Judging whether it is a legal codeword: if it is, output the hard judgment result, otherwise, make a hard judgment again after performing hidden Markov estimation on the information sequence part, and if it is not a legal code sequence, perform the specified number of times LDPC sum-product decoding iterations. Repeat the above joint iteration process until the maximum allowable number of iterations is reached, and output the hard decision result of this decoding. Its information transmission performance is close to Shannong's limit, with small delay and low complexity, and is suitable for applications in wireless multimedia communication.

Description

Message source and channel joint coding method based on the low-density parity test code

Technical field

Message source and channel joint coding method based on low-density checksum coding belongs to communication technical field, the effectively and fast coding method that a kind of fusion hidden Markov information source is estimated and LDPC (low-density checksum) encodes of particularly adopting message source and channel combined coding technology to use in radio multi-media communicating system.

Background technology

In most of radio multi-media communicating systems, source encoding and chnnel coding are to separate consideration, independent design.The foundation of this way is that the mountain farming is theoretical: if rate of information throughput R, then exists a kind of source encoding and a kind of channel coding method less than channel capacity C, can make information reach errorless transmission in noisy channel.Yet in real system, because encoding time delay and encoder complexity are limited, source encoding can not be removed the redundancy of information sequence fully; And chnnel coding also can not reach the error-correcting performance of the mountain farming limit.Particularly in radio multi-media communicating system, the non-redundance compression because source encoding is mainly emphasized frame structure features, and lot of data need and be subjected at limited bandwidth transmitting in the serious channel that disturbs of various bursts, the problem that separating encodes is brought is serious further, has caused the transmission bottleneck of wireless multimedia communication.

Message source and channel combined coding technology is carried out combined optimization with source encoding and chnnel coding; one side can utilize the residual redundancy after the source encoding to help chnnel coding to improve error-correcting performance; on the other hand also can be according to the unequal error protection characteristic of the different code elements of chnnel coding; make remaining error code after the channel decoding to the minimum that influences of source coding, be fit to very much the application in the radio multimedium channel.

Existing message source and channel joint coding method mainly contains two kinds.Method one is traditional method, promptly when carrying out source encoding, consider the characteristic of chnnel coding, making to have certain redundancy through the information sequence after the source encoding, is to consider this part redundancy at channel-decoding, and the error code diffusion that makes residual error code after the channel decoding cause reaches minimum.This method can effectively suppress the limited error code diffusion that brings of chnnel coding code length, but the redundancy application of information source is insufficient, thereby the improvement of performance is not obvious.Method two is the state-of-the-art technology with chnnel coding---the Turbo coding is estimated to combine with the hidden Markov information source, directly carry out the Turbo coding at transmitting terminal to containing redundant multimedia source, the redundancy that the information source of making an uproar is arranged being estimated one as Turbo code becomes demal, the method for employing reaction type serial iteration is carried out joint decoding at receiving terminal.By this method, uniting in the iteration each time, hidden Markov estimates that resulting information source redundancy can improve the error-correcting performance of channel code, and feed back to the input of hidden Markov estimation as next iteration through the information sequence after the channel code error correction, can make and do more accurately, thereby can in next iteration, provide more useful information for channel-decoding to the estimation of information source.This positive feedback process do not needing to know in advance under the prerequisite of information source redundancy properties, can be so that the error-correcting performance of the redundancy of information source and channel Turbo code is fully utilized.Thereby obtained to approach the information transmission performance of the mountain farming limit.But, in the process of associating iterative decoding all be serial algorithm because each composition sign indicating number of Turbo code adopts, need in the iterative process simultaneously to information sequence repeatedly interweave/reciprocal cross knits, make that decoding delay and decoding complexity are excessive, limited the application of this technology in real system.

Summary of the invention

The objective of the invention is to propose a kind of message source and channel combined coding of encoding to solve the deficiency of existing technology based on LDPC.

As shown in Figure 1, at transmitting terminal, contain redundant information source output sequence and need not pass through the information source compressed encoding, directly carry out the LDPC coding, obtain comprising the sign indicating number sequence of information sequence and verification sequence, deliver to next stage then and be modulated into the waveform that is adapted at transmitting in the wireless channel, transmit.At receiving terminal, reception antenna carries out demodulation to the information sequence that receives, and obtains receiving sequence.Adopt the method for combined signal source estimation, channel decoding to decode to receiving sequence then, process is as follows: at first, information sequence in the receiving sequence is partly carried out the hidden Markov parameter Estimation, and according to gained calculation of parameter external information, subsequently receiving sequence and information source are estimated the input of the external information sequence of gained as channel decoding, carry out the LDPC iterative decoding.Through after the iteration of certain number of times, judge whether the Hard decision decoding result of gained is a legal sign indicating number sequence.If then iterative decoding finishes, and exports the information sequence in this hard decision sequence; If be not a legal-code, will decipher revised information sequence through LDPC and feed back to source decoder, carry out once more the hidden Markov parameter Estimation and calculate corresponding external information output.Use the external information sequence corrected received information sequence of gained then, proceed the LDPC iteration.Up to institute's calling sequence is a legal sign indicating number sequence, and perhaps iterations reaches limit value.At this moment, export the hard decision information sequence, begin the joint decoding of next code vector then.

The invention is characterized in that it is the message source and channel joint coding method that a kind of fusion hidden Markov information source is estimated and low-density checksum (LDPC) is encoded, it contains successively and has the following steps:

(1),, carries out the LDPC coding with the LDPC encoder, obtain comprising the sign indicating number sequence of information sequence and verification sequence, transmit delivering to wireless channel after its modulation again containing redundant information source output sequence at transmitting terminal;

(2), at receiving terminal, the information sequence demodulation to receiving obtains receiving sequence;

(3), receiving sequence adopted combined signal source is estimated, the method for channel decoding is decoded, it contains successively and has the following steps:

(3.1), in source decoder, the information sequence in the receiving sequence is carried out the hidden Markov parameter Estimation, and according to gained calculation of parameter external information;

(3.2), again receiving sequence and information source are estimated the input of the external information sequence of gained as channel decoding, in ldpc decoder, carry out and long-pending iterative decoding;

(3.3), through after the iteration of certain number of times, judge whether the Hard decision decoding result of gained is a legal sign indicating number sequence:

If a legal sign indicating number sequence, then iterative decoding finishes, and exports the information sequence in this hard decision sequence;

If not a legal sign indicating number sequence, decipher revised information sequence through LDPC and feed back to source decoder, carry out the hidden Markov parameter Estimation once more and calculate corresponding external information output, use the external information sequence corrected received information sequence of gained then, proceed the LDPC iteration, up to institute's calling sequence is a legal sign indicating number sequence, and perhaps the iterations value of reaching capacity is exported the hard decision information sequence;

(4), if desired, begin the joint decoding of next code vector.

The described combined decoding method of step (3) contains successively when realizing with software and has the following steps:

(a) decoder initialization: the associating number of iterations is changed to zero, and maximum associating iterations is set, and is arranged on simultaneously to unite the LDPC sign indicating number and long-pending decoding iterations in the iteration each time;

(b) receiving sequence is input to decoder;

(c) decoder calculates the hard decision of receiving sequence, judges whether to be a legal-code:

If, export corresponding hard decision result, this decode procedure finishes;

Otherwise, the information sequence in the receiving sequence is partly carried out hidden Markov estimate, and with the external information corrected received sequence of gained; Calculate hard decision, and judge whether the result who obtains is a legal code word through revised receiving sequence:

If, export corresponding hard decision result, this decode procedure finishes;

Otherwise, carrying out the LDPC and the long-pending decoding iteration of stipulated number, the iterations of joint decoding adds 1 when finishing iteration, judges that subsequently the associating iterations is whether less than maximum permissible value:

If, turn back to step (c), continue to carry out corresponding decode procedure;

Otherwise, the hard decision result of calculating and output gained;

(d), judge whether to carry out decode procedure next time according to input condition.

Experiment showed, that it has reached intended purposes.

Description of drawings

Fig. 1. the theory diagram of joint coding method of the present invention.

Fig. 2. the decoding node diagram of duplication code.

Fig. 3. the decoding node diagram of check code.

The decoding grid chart of Fig. 4 .LDPC sign indicating number.

Fig. 5. the present invention unites estimation/decoding schematic diagram.

Fig. 6. software is realized the flow chart of joint decoding algorithm of the present invention.

Embodiment

The principle of the method for the invention and the algorithmic descriptions of employing are as follows:

One. at first, consider when receiving terminal adopts the receiving sequence that contains transmission error, to carry out the algorithm that the hidden Markov information source is estimated.

Considering one is α by the transfering state parameter _IjBinary bits sequence { the u that the two condition Markov information source of (i, j=0,1) produces _k, k=1,2 ..., K}, (K is the length of information bit in the channel code in the formula).Through after the Channel Transmission, the information sequence that contains noise that receives is R _k(k=1,2 ..., K), wherein, R ₁ ⁱExpression is from R ₁To R _iSymbol sebolic addressing.In addition, according to the characteristic of interchannel noise, introduce the transmission error Probability p in original Markov model, making original transition probability is that 1 path becomes with probability 1-p and shifts, and originally transition probability is that 0 path then becomes with Probability p and shifts.Like this, the information sequence through noisy communication channel transmission back gained can adopt this extended model to represent.Under the condition of unknown information source parameter and Channel Transmission error probability, receiving terminal can adopt the hidden Markov estimation approach, obtains this two parameters according to this extended model and receiving sequence, calculates corresponding external information simultaneously.

The initial parameter of supposing this hidden Markov information source is λ={ p{u, s _j| s _i, π, u ∈ 0, and 1}, i, j=0,1}, wherein π is an initial condition, defines s simultaneously _kBe k state of hidden Markov information source, e _k={ s _K-1=i; u _ks _k=j} is a k transfer path, p{u, s _j| s _iBe corresponding transition probability, then the front and back item Iterative Method to this model parameter is:

${\mathrm{\α}}_k\left(i\right)=P\{s_k=i|R_1^k,\mathrm{\λ}\}$

$=\frac{\underset{s_{k-1}}{\mathrm{\Σ}}\underset{u_k}{\mathrm{\Σ}}P\{s_{k-1}=j,u_k,s_k=i,R_1^{k-1},R_k|\mathrm{\λ}\}}{P\left\{R_1^{k-1}\right|\mathrm{\λ}\}\·P\{R_k|R_1^{k-1},\mathrm{\λ}\}}$

$=\frac{\underset{s_{k-1}}{\mathrm{\Σ}}\underset{u_k}{\mathrm{\Σ}}{\mathrm{\α}}_{k-1}\left(j\right)\·P\{u_k,s_k=i|s_{k-1}=j,\mathrm{\λ}\}\·P\{R_k\left|u_k\right\}}{P\left\{R_k\right|R_1^{k-1},\mathrm{\λ}\}}---\left(1\right)$

α in the formula _k(i) be the preceding paragraph equation, and

$P\left\{R_k\right|R_1^{k-1},\mathrm{\λ}\}=\underset{s_{k-1}}{\mathrm{\Σ}}\underset{u_k}{\mathrm{\Σ}}\underset{s_k}{\mathrm{\Σ}}P\{s_{k-1}=j,u_k,s_k=i,R_k|R_1^{k-1},\mathrm{\λ}\}$

$=\underset{s_{k-1}}{\mathrm{\Σ}}\underset{u_k}{\mathrm{\Σ}}\underset{s_k}{\mathrm{\Σ}}{\mathrm{\α}}_{k-1}\left(j\right)\·P\{u_k,s_k=i|S_{k-1}=j,\mathrm{\λ}\}\·P\{R_k\left|u_k\right\}---\left(2\right)$

Equally, consequent equation β _k(i) provide by following formula:

${\mathrm{\β}}_k\left(i\right)=\frac{P\left\{R_{k+1}^K\right|s_k=i,\mathrm{\λ}\}}{P\left\{R_{k+1}^K\right|R_1^k,\mathrm{\λ}\}}$

$=\frac{\underset{u_{k+1}}{\mathrm{\Σ}}\underset{s_{k+1}}{\mathrm{\Σ}}P\{u_{k+1},s_{k-1}=j,R_{k+1},R_{k+2}^K|s_k=i,\mathrm{\λ}\}}{P\left\{R_{k+1}\right|R_1^k,\mathrm{\λ}\}\·P\{R_{k+2}^K|R_1^{k+1},\mathrm{\λ}\}}$

$=\frac{\underset{u_{k+1}}{\mathrm{\Σ}}\underset{s_{k+1}}{\mathrm{\Σ}}P\{u_{k+1},s_{k-1}=j|s_k=i,\mathrm{\λ}\}\·P\{R_{k+1}\left|u_{k+1}\right\}\·{\mathrm{\β}}_{k+1}\left(j\right)}{P\left\{R_{k+1}\right|R_1^k,\mathrm{\λ}\}}---\left(3\right)$

Thereby the probability of k transfer path can be obtained by following formula:

$P\left(e_k\right)=P\{s_{k-1}=i,u_k,s_k=j|R_1^k,\mathrm{\λ}\}$

$=\frac{P\{s_{k-1}=i,u_k,s_k=j,R_1^K|\mathrm{\λ}\}}{P\left\{R_1^K\right|\mathrm{\λ}\}}$

$=\frac{P\{s_{k-1}=i,R_1^{k-1}|\mathrm{\λ}\}}{P\left\{R_1^{k-1}\right|\mathrm{\λ}\}}\·\frac{P\{u_k,s_k=j|s_{k-1}=i,\mathrm{\λ}\}}{P\left\{R_k\right|R_1^{k-1},\mathrm{\λ}\}}$

$\·\frac{P\left\{R_k\right|u_k\}\·P\{R_{k+1}^K|s_k=j,\mathrm{\λ}\}}{P\left\{R_{k+1}^K\right|R_1^k,\mathrm{\λ}\}}$

$=\frac{{\mathrm{\α}}_{k-1}\left(i\right)\·P\{d_k,s_k=j|s_{k-1}=i,\mathrm{\λ}\}\·P\{R_k\left|u_k\right\}\·{\mathrm{\β}}_k\left(j\right)}{P\left\{R_k\right|R_1^{k-1},\mathrm{\λ}\}}---\left(4\right)$

Notice in the transfer path of hidden Markov model any k path e _kAll be by initial state s _i, state of termination s _j, and corresponding information source output u decision.Thereby, transfer path probability P { s _i, u, s _j| λ } can estimate by following formula:

$P\{{\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">s</figure-callout>}}_1,u,{\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">s</figure-callout>}}_1|\mathrm{\&lambda;}\}=\frac{1}{K}\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}P\left(e_k\right|\mathrm{\&lambda;}),{\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">s</figure-callout>}}_1,{\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">s</figure-callout>}}_1\&Element;\{s_0,s_1\},u\&Element;\{\mathrm{0,1}\}---\left(5\right)$

In view of the above, can obtain the amendment type of the probability in each path:

$P^{\text{'}}\{u,s_j|s_i,\mathrm{\&lambda;}\}=\frac{P\{s_i,u,s_j|\mathrm{\&lambda;}\}}{\underset{n}{\mathrm{\&Sigma;}}\underset{s_j}{\mathrm{\&Sigma;}}P\{s_i,u,s_j|\mathrm{\&lambda;}\}}---\left(6\right)$

According to above various, the parameter of hidden Markov information source is modified to λ '={ p ' { u.s by iteration under the noise conditions _j| s _i. λ }. { 0.1}.i.j=0.1}. notices π .u ∈, and when the length long enough of information sequence, the influence that the initial condition error is brought is very little, thereby this algorithm has removed the estimation procedure to the information source initial condition.

By iterative process recited above, the transfer path parameter of hidden Markov model can accurately be estimated.According to these model parameters, the information redundancy that lies in the information source sequence can effectively be extracted, and the form of being write as external information is:

${\mathrm{LLR}}_e\left(u_k\right)=\ln \frac{\underset{s_{k-1}}{\mathrm{\&Sigma;}}\underset{s_k}{\mathrm{\&Sigma;}}{\mathrm{\&alpha;}}_{k-1}\left(i\right)\&CenterDot;P\{u_k=1,s_k=j|s_{k-1}=i,{\mathrm{\&lambda;}}^{\&prime;}\}\&CenterDot;{\mathrm{\&beta;}}_k\left(j\right)}{\underset{s_{k-1}}{\mathrm{\&Sigma;}}\underset{s_k}{\mathrm{\&Sigma;}}{\mathrm{\&alpha;}}_{k-1}\left(i\right)\&CenterDot;P\{u_k=0,s_k=j|s_{k-1}=i,{\mathrm{\&lambda;}}^{\&prime;}\}\&CenterDot;{\mathrm{\&beta;}}_k\left(j\right)}---\left(7\right)$

By these external informations, the transmission error in the receiving sequence can effectively be revised.As an example, table 1 has been listed the bit error rate of hidden Markov estimation front and back received information sequences.The two-state Markov information source that the source bits of emulation this time is 0.81 bit/symbol by an entropy rate produces, and the length of each information sequence is 5000 bits, to each E _b/ N ₀At least carry out the emulation of 1,000,000 bits under (being the average signal power of each bit and the ratio of noise power).By table 1 as seen, the hidden Markov information source estimates that the external information of gained can effectively reduce the mistake that Channel Transmission produces.

The bit error rate of the information sequence before and after table 1. hidden Markov is estimated

E b/N o(dB)	0.0187	0.265	0.466	0.63	0.915
						Bit error rate before hidden Markov is estimated	0.1567	0.1513	0.1465	0.1406	0.1337
Bit error rate after hidden Markov is estimated	0.1295	0.1259	0.1216	0.1160	0.1084

Two. the interpretation method that adopts the LDPC sign indicating number to carry out error control is:

1.LDPC the definition and the parameter of sign indicating number:

The LDPC sign indicating number is a kind of binary packet sign indicating number, and this sign indicating number adopts the supersparsity matrix as check matrix.The number of nonzero element is very rare in every row in the matrix (every row), and the position is random distribution.For convenience of description, the number of nonzero element is the weight of this row (row) in the definition delegation (row).Because the check matrix of LDPC sign indicating number is the matrix that generates at random, the weight of each row (row) is uncertain, therefore adopts the distribution of weight formula to describe this matrix.The column weight amount of same class LDPC code check matrix distributes and can be expressed as with distributed:

$\mathrm{\&lambda;}\left(x\right)=\underset{i=2}{\overset{d_v}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_i{x}^{i-1}---\left(8\right)$

λ in the formula _iExpression weight be i be listed in deal shared in the matrix, d _vValue for the maximum of column weight amount in the matrix.Equally, the capable distribution of weight of same class LDPC code check matrix adopts following formula to describe:

$\mathrm{\&rho;}\left(x\right)=\underset{j=2}{\overset{d_c}{\mathrm{\&Sigma;}}}{\mathrm{\&rho;}}_j{x}^{j-1}---\left(9\right)$

P in the formula _jExpression weight is the row of j shared deal in matrix, d _cMaximum for row weight in the matrix.Because the LDPC sign indicating number is block code, for any legal code word V, with the product of check matrix H be zero, i.e. HV ^T=0.By this check equations as can be known, the same code element of the only corresponding LDPC sign indicating number of the nonzero element of every row in the check matrix has formed a constraint that is equivalent to duplication code.For the ease of the description in the decode procedure, defining this restriction relation is a bit node, and the exponent number of node is the weight of these row.And the nonzero element of every row in the check matrix becomes a constraint that is equivalent to check code with pairing LDPC symbol mapped.It is a check-node that this verification of same definition is closed, and the exponent number of node is the weight of this row.Each nonzero element in the matrix had both participated in the restriction relation of bit node, had participated in the restriction relation of check-node again, thereby can to define the pairing pass of matrix nonzero element be " tie line " that links these two kinds of nodes.In iterative decoding process, decoder utilizes the restriction relation of pairing check-node of the row and column of matrix and bit node to carry out iterative decoding.In iterative process, at first utilize the restriction relation of bit node to decipher, log-likelihood value that is input as the receiving sequence correspondence of each bit node (being probability that each metasymbol is got " 1 " is taken from right logarithm gained again divided by the probability of getting " 0 " value) and relevant check-node are in the output of last once iteration; Subsequently, the output of bit node is delivered to corresponding check-node by " tie line ", utilizes the restriction relation of check-node to decipher again.In this process, a kind of output of node becomes the input of another node, and nonzero element pairing " tie line " becomes " passage " of these two kinds of node input and output exchange messages in the matrix.For code length is the N bit, the LDPC sign indicating number that the column weight amount distributes and the row distribution of weight is determined by (8) (9) two formulas respectively, and the number of its i rank bit node is:

$N_i=N\&CenterDot;\frac{\frac{{\mathrm{\&lambda;}}_i}{i}}{\underset{i=2}{\overset{d_v}{\mathrm{\&Sigma;}}}\frac{{\mathrm{\&lambda;}}_i}{i}}=N\&CenterDot;\frac{\frac{{\mathrm{\&lambda;}}_i}{i}}{{\&Integral;}_0^1\mathrm{\&lambda;}\left(x\right)\mathrm{dx}},2\&le;i\&le;d_v---\left(10\right)$

In like manner, the number of j rank check-node is:

$M_j=M\&CenterDot;\frac{\frac{{\mathrm{\&rho;}}_j}{j}}{\underset{j=2}{\overset{d_c}{\mathrm{\&Sigma;}}}\frac{{\mathrm{\&rho;}}_J}{j}}=M\&CenterDot;\frac{\frac{{\mathrm{\&rho;}}_j}{j}}{{\&Integral;}_0^1\mathrm{\&rho;}\left(x\right)\mathrm{<figure-callout\; id="2"\; label="dx"\; filenames="C0215545900182.png"\; state="\{\{state\}\}">dx</figure-callout>}}2\&le;j\&le;d_c---\left(11\right)$

M is the length of verification code element in the LDPC code word in the formula.

2.LDPC the decoding of sign indicating number:

The supersparsity characteristic of check matrix has fully been used in the decoding of LDPC sign indicating number, calculates and the output external information by the restriction relation of bit node and check-node, and feeds back mutually, carries out iterative decoding.(external information i.e. the information about some code element values that obtains of the restriction relation of all other code elements that belong to a code word by code word, and adopting external information is positive feedback to occur in iterative process alternately.) current, the interpretation method of LDPC sign indicating number is mainly and amasss interpretation method.The log-likelihood value that is input as receiving sequence of this method, and to carrying out iterative decoding by the restriction relation of utilizing bit node and check-node under the number space.At this moment, the restriction relation of bit node show as " with " form, i.e. the output of each bit node be each input log-likelihood value and; Corresponding check-node then shows as the form that certain " amasss ", i.e. the output of each check-node is certain " continued product " of each input log-likelihood value.Because these characteristics, this method is referred to as and amasss interpretation method.

With the information transmission system under the binary input additive white gaussian noise channels is example, and length is weaved into the LDPC code word that length is the N bit for the binary message sequence of N-M bit by the LDPC encoder.Subsequently, this code word is modulated into value and transmits in Gaussian channel for ± 1 symbol sebolic addressing.At receiving terminal, receiver is through sequence of real numbers R that to have obtained a string length that contains noise jamming after the matched filtering be N ₁ ^N, carry out the signal demodulation subsequently.Gaussian channel, BPSK modulation down, i code element be 1 and through modulation and after transmitting receiver receive that signal is R _iProbability be:

$P\left(R_i\right|v_i=1)=\frac{1}{\sqrt{2{\mathrm{\&pi;\&sigma;}}^2}}\exp \{-\frac{1}{{2\mathrm{\&sigma;}}^2}{(R_1-1)}^2\},1\&le;i\&le;N---\left(12\right)$

σ wherein ²Standard variance for interchannel noise.

Equally, i code element be 0 and after modulation and transmission receiver receive that signal is R _iProbability be:

$P\left(R_i\right|v_i=0)=\frac{1}{\sqrt{{2\mathrm{\&pi;\&sigma;}}^2}}\exp \{-\frac{1}{2{\mathrm{\&sigma;}}^2}{(R_i+1)}^2\}1\&le;i\&le;N---\left(13\right)$

By Bayes' theorem, obtain:

$P(v_i=1|R_i)=\frac{P\left(R_i\right|v_i=1)\&CenterDot;P(v_i=1)}{P\left(R_1\right)}---\left(14\right)$

$P(v_i=0|R_i)=\frac{P\left(R_i\right|v_i=0)\&CenterDot;P(v_i=0)}{P\left(R_i\right)}.---\left(15\right)$

In process of transmitting, symbol is got 0 and 1 probability and is equated.For the ease of the output of restituted signal, adopt the form of log-likelihood ratio to represent the maximum a posteriori probability of i code element value receiving usually:

$\mathrm{LLR}\left(R_i\right)=\ln \frac{P(v_i=+1|R_i)}{P(v_i=-1|R_i)}---\left(16\right)$

By above various:

$\mathrm{LLR}\left(R_i\right)=\ln \frac{P(v_i=+1|R_i)}{P(v_i=-1|R_i)}=\ln \frac{P\left(R_i\right|v_i=+1)}{P\left(R_i\right|v_i=-1)}$

$=\mathrm{<figure-callout\; id="1"\; label="ln"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">ln</figure-callout>}\frac{\frac{1}{\sqrt{{2\mathrm{\&pi;\&sigma;}}^2}}\exp \{-\frac{1}{2{\mathrm{\&sigma;}}^2}{(R_i-1)}^2\}}{\frac{1}{\sqrt{{2\mathrm{\&pi;\&sigma;}}^2}}\exp \{-\frac{1}{{2\mathrm{\&sigma;}}^2}{(R_i+1)}^2\}}=\frac{2}{{\mathrm{\&sigma;}}^2}{R}_i$

$=\mathrm{sign}\left(R_i\right)\&CenterDot;\left|\frac{2}{{\mathrm{\&sigma;}}^2}{R}_i\right|---\left(17\right)$

Sign () is a sign function in the formula.In the following formula, first sign function represented the comparative result of the former transmission signal value probability that obtained by received signal.It is 1 probability greater than the probability that is 0 that sign function is got on the occasion of the former symbol of expression: get negative value and represent that then former mother's metasymbol is that 0 probability is greater than the probability that is 1.The size of second absolute value has represented that then this symbol gets the probability of l and get difference degree between 0 the probability.Absolute value is big more, and then the difference of two probable values is big more.Therefore, (17) formula provides two information according to each received signal, and which value is an information get for the original signal most probable, and another information has then been represented the degree of reliability of this judgement.This demodulating process of receiver has fully kept the information of original signal, is called as " soft demodulation ", and perhaps " soft-decision ", corresponding soft-decision output is called " soft information ".

The soft information of demodulator output is sent to ldpc decoder and deciphers.The decoding of ldpc decoder has made full use of the supersparsity characteristic of check matrix, and the restriction relation of check matrix is decomposed into the capable check code restriction relation and the duplication code restriction relation of row, by utilizing the mutual feedback of these two kinds of restriction relations, carries out iterative decoding.For the ease of understanding the decode procedure under these two kinds of restriction relations, the decode procedure of duplication code and check code is discussed at first below.

1) restriction relation of duplication code and decoding thereof:

The coding of duplication code promptly be will input information symbol carry out N-1 repetition, longly be the code word V of N thereby obtain one ₁ ^NTherefore, duplication code has only two legal-codes: complete 0 code word 0 ₁ ^NWith all-ones word 1 ₁ ^NAfter ovennodulation, transmission, demodulation, decoder is deciphered according to the soft information that modulator provides.At the burst that receives is R ₁ ^NPrerequisite under, decipher according to the restriction relation of duplication code, obtain an output sequence U who adopts log-likelihood ratio to represent ₁ ^NWherein, the log-likelihood ratio of i symbol maximum a posteriori probability value is:

$\mathrm{LLR}\left(u_i\right)=\ln \frac{P(v_i=1|R_1^N)}{P(v_i=0|R_1^N)}=\ln \frac{P(v_i=1,R_1^N)}{P(v_i=0,R_1^N)}$

$=\ln \frac{{\mathrm{\&Sigma;}}_{\underset{i^{\&prime;}\&NotEqual;i}{1\&le;i^{\&prime;}\&le;N}}\&CenterDot;\&CenterDot;\&CenterDot;\mathrm{\&Sigma;p}({\mathrm{<figure-callout\; id="1"\; label="v"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">v</figure-callout>}}_1,{\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C0215545900182.png"\; state="\{\{state\}\}">v</figure-callout>}}_2,\&CenterDot;\&CenterDot;\&CenterDot;,v_i=1,\&CenterDot;\&CenterDot;\&CenterDot;,v_N,R_1^N)}{{\mathrm{\&Sigma;}}_{\underset{i^{\&prime;\&NotEqual;i}}{1\&le;i^{\&prime;}\&le;N}}\&CenterDot;\&CenterDot;\&CenterDot;\mathrm{\&Sigma;p}({\mathrm{<figure-callout\; id="1"\; label="v"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">v</figure-callout>}}_1,{\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C0215545900182.png"\; state="\{\{state\}\}">v</figure-callout>}}_2,\&CenterDot;\&CenterDot;\&CenterDot;,v_i=0,\&CenterDot;\&CenterDot;\&CenterDot;,v_N,R_1^N)}$

$=\ln \frac{{\mathrm{\&Sigma;}}_{\underset{i^{\&prime;}\&NotEqual;i}{v_{i^{\&prime;}}\&Element;i_1^{\&prime;N}}}\&CenterDot;\&CenterDot;\&CenterDot;\mathrm{\&Sigma;p}({\mathrm{<figure-callout\; id="1"\; label="v"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">v</figure-callout>}}_1,{\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C0215545900182.png"\; state="\{\{state\}\}">v</figure-callout>}}_2,\&CenterDot;\&CenterDot;\&CenterDot;,v_i=1,\&CenterDot;\&CenterDot;\&CenterDot;,v_N)\&CenterDot;p\left(R_1^N\right|{\mathrm{<figure-callout\; id="1"\; label="v"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">v</figure-callout>}}_1,{\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C0215545900182.png"\; state="\{\{state\}\}">v</figure-callout>}}_2,\&CenterDot;\&CenterDot;\&CenterDot;,v_i=1,\&CenterDot;\&CenterDot;\&CenterDot;,v_N)}{{\mathrm{\&Sigma;}}_{\underset{i^{\&prime;}\&NotEqual;i}{v_{i^{\&prime;}}\&Element;i_1^{\&prime;N}}}\&CenterDot;\&CenterDot;\&CenterDot;\mathrm{\&Sigma;p}({\mathrm{<figure-callout\; id="1"\; label="v"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">v</figure-callout>}}_1,{\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C0215545900182.png"\; state="\{\{state\}\}">v</figure-callout>}}_2,\&CenterDot;\&CenterDot;\&CenterDot;,v_i=0,\&CenterDot;\&CenterDot;\&CenterDot;,v_N)\&CenterDot;p\left(R_1^N\right|{\mathrm{<figure-callout\; id="1"\; label="v"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">v</figure-callout>}}_1,{\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C0215545900182.png"\; state="\{\{state\}\}">v</figure-callout>}}_2,\&CenterDot;\&CenterDot;\&CenterDot;,v_i=0,\&CenterDot;\&CenterDot;\&CenterDot;,v_N)}---\left(18\right)$

Because duplication code has only complete 0 and complete 1 two code words, thereby first in the molecule denominator product term is only when code word is respectively all-ones word and complete 0 code word in the following formula, and probability just is not 0.Thereby (18) formula can continue abbreviation and be:

$\mathrm{LLR}\left(u_i\right)=\ln \frac{p\left(R_1^N\right|V_1^N=1_1^N)}{p\left(R_1^N\right|V_1^N=0_1^N)}=\ln \frac{\underset{i^{\&prime;}=1}{\overset{N}{\mathrm{\&Pi;}}}p\left(R_{i^{\&prime;}}\right|v_{i^{\&prime;}}=1)}{\underset{i^{\&prime;}=1}{\overset{N}{\mathrm{\&Pi;}}}p\left(R_{i^{\&prime;}}\right|v_{i^{\&prime;}}=0)}---\left(19\right)$

$=\underset{i^{\&prime;}}{\overset{N}{\mathrm{\&Sigma;}}}\mathrm{LLR}\left(v_{i^{\&prime;}}\right)=\mathrm{LLR}\left(v_i\right)+\underset{i^{\&prime;}\&NotEqual;i}{\mathrm{\&Sigma;}}\mathrm{LLR}\left(v_{i^{\&prime;}}\right)$

Result's first is the log-likelihood ratio of code element i received signal in the following formula, and the information for code element itself is had is called " prior information "; Second portion is that other code element is called " external information " according to the value information about code element i that the restriction relation of code word provides in the code word.Because prior information just has for each code element itself, thereby decoder only need be given the corresponding external information of each symbol feedback in decode procedure.The decoding relation of duplication code can adopt a node diagram shown in Figure 2 to represent.

Node among Fig. 2 has N bar tie line, a corresponding N code element.These tie lines both can be used as input and also can be used as output, respectively the input of a corresponding N code element and decoding output.In decode procedure, contact is received the demodulating information sequence of representing with the form of log-likelihood ratio by N bar tie line, and subsequently, by the computing of node, decode results is the external information by N code element of these tie lines outputs also.Wherein, every tie line be output as other each bar tie line input value add up and.In the description to the decoding of LDPC sign indicating number, this decoding node of duplication code is also referred to as " bit node ".

2) restriction relation of check code and decoding thereof:

With code check be ^N-1/ _NCheck code be example, the check code that it is N that long information sequence for N-1 bit obtains a code length after through coding, the restriction relation between the code element can be represented with following relational expression:

v ₁v ₂…v _N＝0 (20)

In the formula represent binary system and, i.e. distance in the binary logic.The code word V of check code gained ₁ ^NAfter ovennodulation, transmission, demodulation, obtain comprising a soft information sequence LLR (R of this codeword information ₁ ^N).Check code decoder is promptly deciphered according to this soft information sequence.Define two metasymbol e _iFor in the code word except i code element v _iThe binary system of outer other all code elements and, then by (20) Shi Kede:

v _ie _i＝0 (21)

By binary XOR relation and (21) formula, get code element v _iWith symbol e _iValue is identical.Thereby, code element v _iWith symbol e _iForm a relation that is equivalent to duplication code.The result that obtains is discussed as can be known by top duplication code, code element v _iBy the posterior information of deciphering the back gained be:

$\mathrm{LLR}\left({\hat{v}}_i\right)=\mathrm{LLR}\left(v_i\right)+\mathrm{LLR}\left(e_i\right)---\left(22\right)$

Obviously, second in (22) formula is exactly in decode procedure decoder and feeds back to code element v according to the restriction relation of whole sign indicating number sequence _iExternal information.Below, we are that 3 check code is an example with code length, the expression of derivation external information.Be without loss of generality, we discuss the external information expression of first code element.At code length is under the situation of 3 bits, symbol e ₁Value is that 1 probability is:

p(e ₁＝1)＝p(v ₂＝1)·p(v ₃＝0)+p(v ₂＝0)·p(v ₃＝1)

＝p(v ₂＝1)·(1-p(v ₃＝1))+(1-p(v ₂＝1))·p(v ₃＝1) (23)

＝p(v ₂＝1)+p(v ₃＝1)-2p(v ₂＝1)·p(v ₃＝1)

Thereby,

1-2p(e ₁＝1)＝1-2p(v ₂＝1)-2p(v ₃＝1)+4p(v ₂＝1)·p(v ₃＝1)

＝(1-2p(v ₂＝1))·(1-2p(v ₃＝1)) (24)

Introduce a function:

$\mathrm{\&Phi;}\left(x\right)=\tanh (-\frac{1}{2}x)=\frac{\exp (-\frac{1}{2}x)-\exp \left(\frac{1}{2}x\right)}{\exp (-\frac{1}{2}x)+\exp \left(\frac{1}{2}x\right)}=\frac{1-\exp \left(x\right)}{1+\exp \left(x\right)}---\left(25\right)$

So,

$\mathrm{\&Phi;}\left(\mathrm{LLR}\left(e_1\right)\right)=\mathrm{\&Phi;}\left(\ln \frac{p({\mathrm{<figure-callout\; id="1"\; label="e"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">e</figure-callout>}}_1=1)}{p({\mathrm{<figure-callout\; id="1"\; label="e"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">e</figure-callout>}}_1=0)}\right)=\frac{1-\exp \left(\ln \frac{p({\mathrm{<figure-callout\; id="1"\; label="e"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">e</figure-callout>}}_1=1)}{p({\mathrm{<figure-callout\; id="1"\; label="e"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">e</figure-callout>}}_1=0)}\right)}{1+\exp \left(\ln \frac{p({\mathrm{<figure-callout\; id="1"\; label="e"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">e</figure-callout>}}_1=1)}{p({\mathrm{<figure-callout\; id="1"\; label="e"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">e</figure-callout>}}_1=0)}\right)}=1-2p({\mathrm{<figure-callout\; id="1"\; label="e"\; filenames="C0215545900172.png,C0215545900181.png"\; state="\{\{state\}\}">e</figure-callout>}}_1=1)---\left(26\right)$

By (24), (26) formula, get code element v ₁: the external information representation is:

LLR(e ₁)＝Φ ^-1(Φ(LLR(v ₂)·Φ(LLR(v ₃)) (27)

(27) formula can be generalized to any one code element, also can be generalized to the situation of code length greater than 3 bits.Get under the situation for the N bit at code length, the external information of code element i is:

$\mathrm{LLR}\left(e_i\right)={\mathrm{\&Phi;}}^{-1}\left(\underset{i^{\&prime;}\&NotEqual;i}{\underset{1\&le;i^{\&prime;}\&le;N}{\mathrm{\&Pi;}}}\mathrm{\&Phi;}\left(\mathrm{LLR}\left(v_i\right)\right)\right)---\left(28\right)$

This decoding operation relation of check code also can adopt a node to represent, as shown in Figure 3:

Node has N tie line among Fig. 3, a corresponding N code element; Every tie line be input also be output.Wherein, input is corresponding to the soft information sequence that is input to decoder, and output then is decoder feeds back to each symbol by computing external information.In once deciphering, every tie line is imported the soft information that obtains after this symbol demodulation and is arrived node, by computing, gives each tie line an external information output with posterior nodal point.Notice that the output of every tie line is with the input value of other all tie lines operation result as input among Fig. 3.In the decoding of follow-up LDPC sign indicating number, this decoding node of check code is called as " check-node ".

3) concrete decode procedure:

After the coding of information sequence process LDPC sign indicating number, modulation, the transmission, carry out matched filtering, comprised the receiving sequence R of LDPC codeword information accordingly by receiver ₁ ^N, this sequence is sent to ldpc code decoder and carries out error-correcting decoding subsequently.In decode procedure, decoder at first carries out demodulation to receiving sequence, receiving sequence is converted into the form of soft information; Subsequently, utilize the check equations HV of LDPC sign indicating number ^T=0 deciphers.The check matrix of noticing the LDPC sign indicating number is the supersparsity matrix, and the nonzero element number of every row/row is very rare.Know by check equations, the product of the every capable LDPC sign indicating number of matrix, be actually the code element that multiplies each other with this row nonzero element binary system and.By the constraint equation of check code as can be known, these code elements have constituted the constraint of a check code.Because check matrix has M capable, thereby can obtain M check code altogether.By adopting the interpretation method of check code, each check code can be given the external information output of this code element value condition of a reflection of each code element under restriction relation separately.And for each row of check matrix because its element only multiplies each other with same code element in the multiplication of check matrix and code word, and each nonzero element that should row all check code of correspondence to the output of this symbol value condition.So the output of these check codes has constituted the constraint of a duplication code with the soft information of the code element that receives.Since the total N row of check matrix, thereby can obtain N duplication code, corresponding with N code element of LDPC code word respectively.The decoding of LDPC sign indicating number promptly is the restriction relation that is decomposed into this M check code and N duplication code by the restriction relation with check matrix, exports the input that is fed back to the other side mutually by the decoding of these two kinds of sign indicating numbers, carries out parallel iteration decoding.By above discussion about duplication code and check code, the decoding grid chart of LDPC sign indicating number can be represented by Fig. 4:

At first, after the decoded device of receiving sequence was converted into soft information, decoder was made as 0 with the initial output of all check-nodes, decoding when carrying out N bit node according to the initial output of the soft information of receiving sequence and check-node subsequently.These bit nodes are to the external information output of each code element, delivered to corresponding check-node by tie line, M node carries out the decoding of check code simultaneously subsequently, and the decoding output to each symbol of each check-node all feeds back to relevant bit node by tie line.When next iteration began, each bit node all added up own all inputs, obtains the posterior information of a code element, carries out Hard decision decoding according to this posterior information subsequently.The Hard decision decoding of N bit node obtains the valuation information sequence of a code word.If the product of check matrix and this valuation information sequence is zero, then decoder stops iterative decoding and exports this valuation as decode results; Otherwise decoder carries out the decoding iteration of bit node-check-node next time, up to gained valuation sequence be a legal LDPC code word or reach maximum iteration time till.Decoder is output as the hard decision valuation sequence that obtains for the last time.

If r _IjFor output to the external information of bit node i, q from check-node j _IjBe external information from bit node i to check-node j, should and the iterative process of long-pending interpretation method comprise the steps:

1) decoding initialization: for the sequence of real numbers R that receives ₁ ^N, the initial reception of corresponding i code element of LDPC sign indicating number is worth the form that decoded device is demodulated to log-likelihood ratio:

$\mathrm{LLR}\left(R_i\right)=\frac{2}{{\mathrm{\&sigma;}}^2}{R}_i,1\&le;i\&le;N---\left(29\right)$

LLR represents that value is a log-likelihood ratio in the formula, σ ²Standard variance for interchannel noise.Simultaneously, check-node without any the information about code word so the external information that check-node j outputs to bit node i is set is under the initial condition:

LLR(r _ij)＝0 (30)

2) if the hard decision result of resulting sequence is not that (wherein hard decision is meant that log-likelihood value according to each symbol of sequence determines the bit value of each symbol to a legal code word, the log-likelihood value be positive number then code element get symbol " 1 ", for negative then code element get symbol " 0 "), carry out once and the iterative process of long-pending decoding is:

A) decoding of bit node: under the restriction relation of this node, output with the input the pass be " with " relation, promptly bit node i is output as to the external information of check-node j:

$\mathrm{LLR}\left(q_{\mathrm{ij}}\right)=\underset{\underset{j^{\&prime;}\&NotEqual;j}{j^{\&prime;}\&Element;\mathrm{Col}\left[i\right]}}{\mathrm{\&Sigma;}}\mathrm{LLR}\left(r_{\mathrm{ij}}\right)+\mathrm{LLR}\left(R_i\right)---\left(31\right)$

Col[i in the formula] location sets of expression check matrix H i row nonzero element.

B) decoding of check-node: under the restriction relation of check-node, output is certain relation of " amassing " with the pass of input, and promptly check-node j outputs to the external information of bit node i and is:

$\mathrm{LLR}\left(r_{\mathrm{ij}}\right)={\mathrm{\&Phi;}}^{-1}\left(\underset{i^{\&prime;}\&NotEqual;i}{\underset{i^{\&prime;}\&Element;\mathrm{Row}\left[j\right]}{\mathrm{\&Pi;}}}\mathrm{\&Phi;}\left(\mathrm{LLR}\left(q_{i^{\&prime;}j}\right)\right)\right)---\left(32\right)$

Row[j in the formula] location sets of expression check matrix H j capable nonzero element, and

$\mathrm{\&Phi;}\left(x\right)=\tanh (-\frac{1}{2}x)---\left(33\right)$

3) after the iteration decode results of i bit node of gained be these all inputs of node and:

$\mathrm{LLR}\left({\hat{v}}_i\right)=\underset{j^{\&prime;}\&Element;\mathrm{Col}\left[i\right]}{\mathrm{\&Sigma;}}\mathrm{LLR}\left(r_{{\mathrm{ij}}^{\&prime;}}\right)+\mathrm{LLR}\left(R_i\right)---\left(34\right)$

Resulting decode results is carried out following hard decision, transferred to for second step then.Wherein the hard decision of i symbol is:

${\hat{u}}_i=\left\{\begin{array}{ccc}1& \mathrm{if}& \mathrm{LLR}\left({\hat{v}}_i\right)>0\\ 0& \mathrm{if}& \mathrm{LLR}\left({\hat{v}}_i\right)<0\end{array}---\left(35\right)\right.$

4) carry out the decoding of next code word if desired, jump to the first step; Otherwise, finish decoding.

This interpretation method has made full use of the character of bit node and check-node in the supersparsity matrix, and the soft information of receiving sequence, thereby can obtain approaching the decoding performance of the mountain farming limit with very low decoding complexity.Simultaneously, this algorithm is a parallel algorithm, can effectively reduce decoding delay.

Three. in conjunction with above two algorithms, the present invention proposes and merge the combined signal source channel interpretation method that hidden Markov is estimated and LDPC encodes.

The present invention is based on the random code characteristic of LDPC sign indicating number, removed interleaver, directly information source sequence and chnnel coding are cascaded up at transmitting terminal.At the decoding end, the structure chart of this joint decoding as shown in Figure 5.The first half is bit one check-node of LDPC sign indicating number among the figure, and the node E of lower part represents the hidden Markov of information source information bit sequence is estimated.In iterative decoding algorithm of the present invention, to receiving sequence R ₁ ^KHidden Markov estimate at first to carry out, utilize the external information sequence { LLR of gained _e(u _i), 1≤i≤K} revises receiving sequence:

$\left\{\begin{array}{cc}\mathrm{LLR}\left(p_i\right)=\frac{2}{{\mathrm{\&sigma;}}^2}{R}_i+{\mathrm{LLR}}_e\left(u_i\right)& 1\&le;i\&le;K\\ \mathrm{LLR}\left(p_i\right)=\frac{2}{{\mathrm{\&sigma;}}^2}{R}_i,& K+1\&le;i\&le;N\end{array}---\left(36\right)\right.$

Subsequently, will deliver to channel decoder through revised sequence, at bit node { v _i, 1≤i≤N} and check-node { c _j, carry out the LDPC iterative decoding between 1≤j≤M}.As shown in Figure 5, the first half is the bit-check-node of LDPC sign indicating number among the figure, and the node E of lower part represents the hidden Markov of information source information bit sequence is estimated.Whole joint decoding process may further comprise the steps:

At first, estimate node 5c, the hidden Markov of received information sequence is estimated at first to carry out, subsequently with the external information sequence { LLR of gained at hidden Markov _e(u _i), 1≤i≤K} outputs to bit node 5a.At bit node 5a, receiving sequence is revised according to (36) formula, carry out initialization according to (30) formula simultaneously.Then, calculating hard decision that each bit joint 5a order according to (34) (35) formula exports and judges whether to be a legal LDPC code word.If a legal code word, then decoding finishes, and exports corresponding hard decision result; Otherwise, carry out one time iterative process: the output that bit node 5a calculates each node according to (31) formula, deliver to corresponding check-node 5b as input by line 5d between node; Check-node 5b calculates the external information that feeds back to each bit node 5a according to (32) formula again, and its input as bit node 5a next iteration.After finishing these computings, iterations adds 1.When next iteration began, each bit node 5a calculates hard decision output according to (34) (35) formula once more and whether judgement is a legal-code.If a legal-code then finishes the decoding iteration, export corresponding hard decision sequence; Otherwise, carry out iterative process once more one time.After decoding iteration certain number of times and long-pending, if the hard decision series that obtains is not a legal-code, then will feed back to source decoder 5c through the information sequence of channel decoding gained external information correction, carrying out hidden Markov estimates, subsequently, the external information sequence of gained is outputed to bit node 5a, proceed LDPC decoding iteration.Reach maximum up to obtaining a legal-code or iterations.In this case, the hard decision information sequence of output gained is as decode results.Whole joint decoding process can be summarized as follows:

1) according to receiving sequence R ₁ ^K, carry out once hidden Markov estimation procedure as described above, the external information of gained is sent to corresponding information bit node by line;

2) employing and long-pending decoding algorithm carry out the decoding iteration of the LDPC sign indicating number of certain number of times (can select 5,10,20 times according to the needs of Practical Calculation complexity and error-correcting performance), calculate corresponding hard decision output then;

3) if step 2) output of the hard decision of gained is not a legal LDPC code word, and the iterations that iterations allows less than maximum, resulting external information of channel decoding and received information sequence are fed back to source decoder, carry out hidden Markov iteration one LDPC iteration joint decoding process again one time; Otherwise, release the joint decoding process and export the hard decision sequence of gained.

Notice, in algorithm of the present invention, the redundancy of information source is estimated to carry out prior to the decoding of LDPC sign indicating number, obtain certain useful information like this, reduce the erroneous level in the receiving sequence to a certain extent.Thereby the iterative decoding of LDPC sign indicating number is converged on the correct code word faster.In addition, the present invention adopted hidden Markov iteration and repeatedly LDPC and long-pending iterative decoding combine unite estimation/interpretation method, the powerful error-correcting performance of LDPC sign indicating number is given full play to, can also make follow-up each information source estimate to do accurate and effective more, thereby can provide more useful information to channel decoding.Simultaneously, this method also effectively reduces decoding complexity and decoding delay.

Below, come more detailed explanation the present invention with embodiment with reference to the accompanying drawings:

Embodiment one: present embodiment is for realizing the system flow of the joint coding method that the present invention proposes, as shown in Figure 1, may further comprise the steps: at transmitting terminal, contain redundant information source 11 output sequences 12, need not pass through the information source compressed encoding, and directly carry out chnnel coding, obtain comprising the sign indicating number sequence 14 of information sequence and verification sequence by LDPC encoder 13, deliver to next stage then and be modulated into the waveform that is adapted at transmission in the wireless channel 15, transmit.At receiving terminal, reception antenna carries out demodulation to the information sequence that receives, and obtains receiving sequence 16.Adopt the method for combined signal source estimation, channel decoding to decode to receiving sequence then, process is as follows: at first, in source decoder 18, the information sequence in the receiving sequence 16 17 is carried out the hidden Markov parameter Estimation, and according to gained calculation of parameter external information, subsequently receiving sequence 16 and information source are estimated the input of the external information sequence 19 of gained as channel decoding, in ldpc decoder 1a, carry out and long-pending iterative decoding.Through after the iteration of certain number of times, judge whether the Hard decision decoding result of gained is a legal sign indicating number sequence.If a legal sign indicating number sequence, then iterative decoding finishes, and exports the information sequence 1b in this hard decision sequence; If be not a legal-code, will decipher revised information sequence 1c through LDPC and feed back to source decoder 18, carry out once more the hidden Markov parameter Estimation and calculate corresponding external information output.Use the external information sequence corrected received information sequence of gained then, proceed the LDPC iteration.Up to institute's calling sequence is a legal sign indicating number sequence, and perhaps iterations reaches limit value.At this moment, export hard decision information sequence 1b, begin the joint decoding of next code vector then.

Embodiment two: present embodiment is realized the general flow of the combined decoding method that the present invention proposes for adopting software, as shown in Figure 6, during the decoding beginning, decoder forwards 6b to from step 6a, carry out initialization: the associating number of iterations is changed to 0, maximum associating iterations is set, is arranged on simultaneously and unites the LDPC sign indicating number and long-pending decoding iterations in the iteration each time.Subsequently, at step 6c, receiving sequence is input to decoder; At step 6d, decoder calculates the hard decision of receiving serial, and judges whether to be a legal-code.If, then jump to step 6i, export corresponding hard decision result, finish this decode procedure subsequently; Otherwise decoder forwards step 6e to, the information sequence in the receiving sequence is partly carried out hidden Markov estimate, and with the external information corrected received sequence of gained.After finishing these computings, decoder forwards step 6f to, calculates the hard decision through revised receiving sequence, and judges whether the result of gained is a legal-code.If, then jump to step 6i, export corresponding hard decision result, finish this decode procedure subsequently; Otherwise decoder forwards step 6g to, carries out the LDPC and the long-pending decoding iteration of specific times, and the iterations of joint decoding adds 1 when finishing iteration.Subsequently, decoder judges that at step 6h whether the associating iterations is less than maximum permissible value.If, then jump to step 6d, continue to carry out corresponding decode procedure; Otherwise decoder forwards step 6i to, calculates and export the hard decision result of gained.At step 6j, decoder judges whether that according to input condition needs carry out decode procedure next time.If, then jump to step 6c, begin the decoding of next receiving sequence; Otherwise decoder finishes the decoding state.

Should be pointed out that method of the present invention can also promote the use of in other broadband cabled communication system such as High Data Rates such as high definition cable tv broadcast system, high reliability goes.

Effect of the present invention is, estimates and the combination of LDPC code by hidden Markov model, obtained with method two equally near the information transmission performance of the agricultural limit in mountain; Simultaneously, save interleaving process in the combined decoding by the random code characteristic of utilizing the LDPC code, and utilize the parallel characteristics of the iterative decoding algorithm of LDPC code, can obtain ratio method two much smaller decoding complexity and decoding delay. Therefore, compare with existing method, joint coding method of the present invention is more suitable for the application in radio multi-media communicating system.

Claims (2)

1. The source-channel joint coding method based on low-density parity-check coding, which contains the step of hidden Markov source estimation, is characterized in that it is a fusion of hidden Markov source estimation and low-density parity check Coded source-channel joint coding method, low-density parity check is also called LDPC, and the method contains the following steps in turn: (1), at the sending end, carry out LDPC encoding to the source output sequence containing redundancy with an LDPC encoder to obtain a code sequence including an information sequence and a check sequence, and then modulate it and send it to a wireless channel for transmission; (2), at the receiving end, demodulate the received information sequence to obtain the received sequence; (3), the received sequence is decoded by joint source estimation and channel decoding, which contains the following steps in turn: (3.1), perform hidden Markov parameter estimation on the information sequence in the received sequence in the source decoder, and calculate the external information according to the obtained parameters; (3.2), then use the received sequence and the external information sequence obtained by source estimation as the input of channel decoding, and perform sum-product iterative decoding in the LDPC decoder; (3.3), after a certain number of iterations, judge whether the obtained hard decision decoding result is a legal code sequence: If it is a legal code sequence, the iterative decoding ends, and the information sequence in the hard decision sequence is output; If it is not a legal code sequence, the information sequence corrected by LDPC decoding is fed back to the source decoder, the hidden Markov parameter is estimated again and the corresponding external information output is calculated, and then the received external information sequence is used to correct the received code sequence. Information sequence, continue to perform LDPC iterations until the obtained sequence is a legal code sequence, or the number of iterations reaches the limit value, and output a hard decision information sequence; (4) If necessary, start the joint decoding of the next code vector. 2. the source-channel joint coding method based on low-density parity-check coding according to claim 1, is characterized in that, the joint decoding method described in step (3) contains following steps successively when realizing with software: (d) Decoder initialization: the number of joint iterations is set to zero, the maximum number of joint iterations is set, and the number of sum-product decoding iterations of the LDPC code in each joint iteration is set at the same time; (e) The received sequence is input to the decoder; (f) The decoder calculates the hard decision of the received sequence to determine whether it is a legal codeword: If yes, output the corresponding hard decision result, and the decoding process ends; Otherwise, perform a hidden Markov estimation on the information sequence part in the received sequence, and use the obtained external information to modify the received sequence; calculate the hard decision of the corrected received sequence, and judge whether the obtained result is a legal codeword : If yes, output the corresponding hard decision result, and the decoding process ends; Otherwise, perform a specified number of LDPC sum-product decoding iterations, add 1 to the number of iterations of joint decoding at the end of the iteration, and then determine whether the number of joint iterations is less than the maximum allowable value: If yes, return to step (3), and continue to perform the corresponding decoding process; Otherwise, calculate and output the resulting hard decision result; (4) According to the input situation, it is judged whether the next decoding process needs to be performed.

Images