CN107241104A - One local contrary sign dynamic BP interpretation method for LDPC code - Google Patents
One local contrary sign dynamic BP interpretation method for LDPC code Download PDFInfo
- Publication number
- CN107241104A CN107241104A CN201710270488.3A CN201710270488A CN107241104A CN 107241104 A CN107241104 A CN 107241104A CN 201710270488 A CN201710270488 A CN 201710270488A CN 107241104 A CN107241104 A CN 107241104A
- Authority
- CN
- China
- Prior art keywords
- node
- check
- residual error
- variable node
- message
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/03—Error detection or forward error correction by redundancy in data representation, i.e. code words containing more digits than the source words
- H03M13/05—Error detection or forward error correction by redundancy in data representation, i.e. code words containing more digits than the source words using block codes, i.e. a predetermined number of check bits joined to a predetermined number of information bits
- H03M13/11—Error detection or forward error correction by redundancy in data representation, i.e. code words containing more digits than the source words using block codes, i.e. a predetermined number of check bits joined to a predetermined number of information bits using multiple parity bits
- H03M13/1102—Codes on graphs and decoding on graphs, e.g. low-density parity check [LDPC] codes
- H03M13/1105—Decoding
- H03M13/1111—Soft-decision decoding, e.g. by means of message passing or belief propagation algorithms
- H03M13/1125—Soft-decision decoding, e.g. by means of message passing or belief propagation algorithms using different domains for check node and bit node processing, wherein the different domains include probabilities, likelihood ratios, likelihood differences, log-likelihood ratios or log-likelihood difference pairs
Landscapes
- Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Error Detection And Correction (AREA)
Abstract
The present invention is directed to the dynamic BP decoding algorithm of LDPC code, propose a kind of local dynamic station based on LDPC code and update interpretation method (LILRBP), this method uses the information updating based on local residual error, effectively improve including the BP decoding performances in a small amount of iterations and under high s/n ratio, surmounted other dynamic BP decoding algorithms.From unlike conventional dynamic BP interpretation method, it is ageing that LILRBP methods think that residual error has, therefore time upper this nearest part residual error is only focused on, and set iterations threshold value, under threshold value, local residual error is screened again according to correlated variables node likelihood ratio sign change situation, next message to be updated then is determined by the residual error filtered out;Next message to be updated directly then is determined by local residual error on threshold value.Advantage of this is that total energy uses newest message to carry out information updating, play a part of accelerating algorithm convergence rate, and local residual error is screened again under threshold value, even more further speed up algorithmic statement, both combine joint effect convergence, the convergence rate of BP decoding algorithms is effectively improved, the purpose of lifting decoding performance is reached.
Description
Technical field
The present invention relates to a kind of LDPC code decoding technique field, more particularly to a kind of local contrary sign dynamic based on LDPC code
Interpretation method.
Background technology
Since LDPC code was found again from 1996, its decoding algorithm (flooding BP algorithms) is simple with its realization,
The features such as decoding complexity is linearly increasing, becomes coding and decoding field focus.And SVNF-RBP algorithms are translated for the dynamic BP of representative
Although code algorithm adds residual computations and search complexity, while also greatly improving the decoding performance of BP algorithm.
SVNF-RBP decoding algorithms are an asynchronous dynamical message iterative algorithms, each time according to a maximum verification section
Point arrives variable node message residual error, positions next check-node to be updated to variable node message.Information updating is exactly root
It is sequence according to the maximum residul difference found every time, is transmitted back and forth along the side in the Tanner figures corresponding with LDPC code, its message is passed
Main two steps of lateral calculations and longitudinal direction calculating included for each variable node are passed, wherein lateral calculations are exactly school
Test node ciTo variable node vjMessage transmission:
It is exactly variable node v that longitudinal direction, which is calculated,iTo check-node cjMessage transmission:
It is finally that 0,1 judgement is done by the maximum likelihood ratio of each variable node in BP decoding algorithms.Each
Variable node will all receive the prior probability for carrying out self-channel(pv(0), pv(1) represent to pass respectively
The bit passed is 0 and 1 probability), also to receive the message from each check-node transmission being attached thereto.Therefore variable
Node viLikelihood ratio be exactly all message summations received:
Likelihood ratio message residual computations formula:r(mk)=| | fk(m)-mk||∞, mk∈m;Wherein m represents to calculate fk(m) institute
The related news needed, mkAnd fk(m) likelihood ratio before check-node updates to variable node and after renewal is represented respectively.
SVNF-RBP algorithm iterations process stops in a period of time for meeting following condition:
(1) all check equations are all met.
(2) iterations reaches the maximum of setting.
The detailed process of SVNF-RBP algorithms is:
1) all m are initializedc,v=0;
2) initialize all
3) all r (m are calculatedc,v);
4) to each vj, find out
5) calculateWithca∈N(vmax)\cmax
6) calculateca∈N(vmax)\cmax,vb∈N(ca)\vmax
If 7) all check equations meet or reached the maximum iteration of setting, terminate decoding, otherwise return
Step 4)
For BP iterative algorithms, asynchronous strategy is typically all by accelerating decoding convergence to lift decoding performance.SVNF-
BP decoding performances can be substantially improved in RBP algorithms, but are the increase in substantial amounts of calculating, search complexity.Therefore, in lifting decoding property
Reduction complexity is particularly important while energy.
The content of the invention
It is an object of the invention to overcome at least one shortcoming and deficiency of above-mentioned prior art there is provided one kind to be based on LDPC
The local dynamic station interpretation method of code, the local contrary sign dynamic decoding method effectively improves the decoding under especially high s/n ratio
Can, while reducing residual error search and storage complexity.
The purpose of the present invention is achieved through the following technical solutions:
A kind of dynamic BP interpretation method based on LDPC code, proposes that residual error has ageing, and sets iterations threshold value,
In the range of the local residual error of newest generation, when iterations is less than threshold value, with reference to the likelihood ratio symbol of correlated variables node
Change is screened to local residual error again, and the update sequence that maximum residul difference determines message is found out in the residual error filtered out;When repeatedly
When generation number is higher than threshold value, the update sequence that the maximum determines message is found out in local residual error.
The check-node to be updated is selected to variable node messageNew information firstTo all verification sections
Point ca∈N(vj)\ciProduce and transmit messageTo all variable node vb∈N(ca)\vjCalculate message residual error
When iterations is less than iterations threshold value IthrWhen, allIn find out variable node vbLikelihood ratio symbol will
The residual error changed, and find out the maximum wherein and determine next check-node to be updated to variable node message;When
Iterations is more than iterations threshold value IthrWhen, allIn find out the maximum and determine next verification to be updated
Node is to variable node message.
Local contrary sign dynamic decoding algorithm (LILRBP):
1) all m are initializedc,v=0
2) initialize all
3) all r (m are calculatedc,v)
4) in all r (mc,v) in find out the related residual for the v that likelihood ratio symbol can change, and find out wherein
5) calculateWithca∈N(vmax)\cmax
6) calculateca∈N(vmax)\cmax,vb∈N(ca)\vmax
7) I < IthrWhen, allIn find out the v that likelihood ratio symbol can changebRelated residual, and
Find out whereinI > IthrWhen, allIn find out
If 8) all check equations meet or reached the maximum iteration of setting, terminate decoding, otherwise return
Step 5)
Wherein:mc,vRefer to all check-nodes to the message of variable node;Refer to variable node vnIt is connected to all
Check-node message;Represent variable node viChannel prior probability;r(mc,v) refer to all check-nodes to variable
Residual error before the information updating of node and after renewal;Represent check-node ciTo variable node vjLikelihood ratio update
Residual error after preceding and renewal;N(vi) represent and variable node viConnected all check-node set, check-node ca∈N(vi)
Represent check-node caTo take all over variable node viBe connected all check-nodes;Represent from check-node caTo variable section
Point viMessage,Represent from variable node viTo check-node cjMessage;N(cj) represent and check-node cjConnected institute
There are variable node, N (cj)\viRepresent and check-node cjConnected is all not including variable node viOther variable nodes, vb
∈N(cj)\viRepresent variable node vbTo take all over check-node cjConnected is all not including variable node viIts dependent variable
Node;I represents iterations, and IthrIt is iterations threshold value.
Likelihood ratio message residual computations formula:r(mk)=| | fk(m)-mk||∞, mk∈m;Wherein m represents to calculate fk(m) institute
The related news needed, mkAnd fk(m) likelihood ratio before check-node updates to variable node and after renewal is represented respectively.
It is ageing that this method proposes that the generation of residual error has, and sets iterations threshold value, using the threshold value as boundary, iteration time
When number is less than threshold value, screened in local residual error scope in conjunction with the likelihood ratio sign change situation of correlated variables node,
And find out maximum residul difference to set up a check-node to the renewal sequence of variable node message in these residual errors selected;Repeatedly
When generation number is higher than threshold value, maximum residul difference is found out in local residual error to set up a check-node to variable node message more
New sequence,
Therefore, the present invention has the following advantages and effect relative to prior art:
The residual error of maximum is only found out from the residual error of newest generation, can be scanned for while calculating, therefore residual error is not
Storage is needed, storage complexity is reduced;The namely local residual error of the residual error of newest generation, reduces answering for search maximum residul difference
Miscellaneous degree;Less than iterations threshold value, local residual error always can ensure that the information updating that next step is carried out using latest news, and office
The residual error that portion's residual error combination variable node likelihood ratio sign change situation is filtered out again, can further accelerate variable node and turn over
Turn, reach the convergence rate for accelerating decoding process.
Relative to SVNF-RBP algorithms, the present invention can not only improve the decoding performance of less iterations, more greatly
Improve the decoding performance of high s/n ratio.
Brief description of the drawings
Fig. 1 is algorithm LILRBP dynamic strategy schematic diagram.
Fig. 2 is algorithm flooding, LBP, and NW RBP, SVNF-RBP and LILRBP are in code length 1944, code checkGiven letter
The FER performance maps made an uproar during than for 1.75dB.
Fig. 3 is algorithm SVNF-RBP and LILRBP in code length 1944, code check5th, FER performance maps during 50 iteration.
Embodiment
With reference to embodiment and accompanying drawing, the present invention is described in further detail, but embodiments of the present invention are not limited
In this.
The present invention is a kind of local dynamic station interpretation method based on LDPC code, is the local contrary sign dynamic BP for LDPC code
Interpretation method (LILRBP), it is ageing that this method proposes that the generation of residual error has, and only focuses on the local residual error of nearest time generation,
And iterations threshold value is set, when iterations is less than threshold value, the change of correlated variables node likelihood ratio is combined in local residual error
Situation screening residual error is changed, and finds out maximum residul difference to set up a check-node to variable node message in the residual error filtered out
Renewal sequence;When iterations is higher than threshold value, directly find out maximum residul difference in local residual error to set up a check-node
To the renewal sequence of variable node message.
If I represents iterations, and IthrIt is iterations threshold value;N(vi) represent and variable node viConnected all schools
Test node, N (vi)\cjThen represent and variable node viConnected does not include check-node cjEvery other check-node;N(ci)
Represent and check-node (check equations) ciConnected all variable nodes, N (ci)\vjThen represent and check-node ciConnected
Do not include variable node vjEvery other variable node.The variable node and check-node Message function definable of interconnection
ForWherein m represents to calculateOrRequired related news.
The prior probability of channel(pv(0), pv(1) bit information for representing transmission respectively is 0 and 1
Probability).Check-node is to variable node likelihood ratio message residual computations formula r (mk)=| | fk(m)-mk||∞, wherein m represents
Calculate fk(m) related news needed for, mkAnd fk(m) represent respectively before check-node updates to variable node message and after renewal
Likelihood ratio.
Assuming that check-node is to variable node messageWith local maxima residual error, the dynamic strategy bag of the decoding algorithm
Include following three step:
First, new informationAnd transmit.
Secondly, new informationca∈N(vj)\ciAnd transmit.
Finally, residual error is calculatedvb∈N(ca)\vj,ca∈N(vj)\ci, as I < IthrWhen, all
In find out the v that likelihood ratio symbol can changebRelated residual, and find out maximum residul difference wherein;I > IthrWhen, institute
HaveIn find out maximum residul difference.
Specifically, algorithm iteration process of the invention is as follows:
1) all m are initializedc,v=0
2) initialize all
3) all r (m are calculatedc,v)
4) in all r (mc,v) in find out the related residual for the v that likelihood ratio symbol can change, and find out wherein
5) calculateWithca∈N(vmax)\cmax
6) calculateca∈N(vmax)\cmax,vb∈N(ca)\vmax
7) I < IthrWhen, allIn find out the v that likelihood ratio symbol can changebRelated residual, and
Find out whereinI > IthrWhen, allIn find out
If 8) all check equations meet or reached the maximum iteration of setting, terminate decoding, otherwise return
Step 5)
During an iteration of dynamic BP decoding algorithm, otherwise check-node to variable node message amount of calculation with
BP algorithm is identical, or the message amount of calculation of variable node to check-node is identical with BP algorithm, all emulation all will be strict
In accordance with this rule.Following table provides the message amount of calculation to this algorithm an iteration process, wherein,WithRepresent to become respectively
The average degree of node and check-node is measured, e represents the quantity on side in Tanner figures, had simultaneouslyInto
Vertical, wherein N and M represent the number of variable node and check-node in Tanner figures respectively.
Check-node in an iteration of table 1 is to variable node message amount of calculation
Variable node in an iteration of table 2 is to check-node message amount of calculation
In table 1,2, Flooding BP algorithms are used as asynchronous non-dynamic message as synchronization message more new algorithm, LBP algorithms
More new algorithm, is only referred to.
Dynamic algorithm NW RBP algorithms and SVNF-RBP algorithms are all residual to variable node message maximum using check-node
Difference, maximum residul difference calculation formula:r(mk)=| | fk(m)-mk||∞, wherein r (mk) maximum message residual error is represented, m represents to calculate fk
(m) related news needed for, mkAnd fk(m) likelihood ratio before check-node updates to variable node and after renewal is represented respectively, is come
Next check-node to be updated is found out to variable node message, then deploys the information updating process of an essence, two
Algorithm is required for storing residual error, except that NW RBP algorithms are to disappear in whole non-zero check-nodes to variable node
Next message preferentially to update is chosen in breath residual error, and SVNF-RBP algorithms are in each variable node coverage
Non-zero check-node next message preferentially to update is chosen into variable node message residual error.
Related algorithm principal character is given below:
NW RBP algorithms:
1) all m, are initializedc,v=0;
2), initialize all
3) all r (m, are calculatedE, v);
4), maximum residul difference is selected in all non-zero residual errors
5), to each vk∈N(ci), calculateWhile handleIt is set to 0.
6), to each ca∈N(vk), calculate
7), to each vb∈N(ca)\vk, calculate
If 8), all check equations meet or reached the maximum iteration of setting, then terminate decoding, otherwise return
Return step 4.
SVNF-RBP algorithms:
1) all m, are initializedc,v=0;
2), initialize all
3) all r (m, are calculatedE, v);
4), to each vi, to all vb∈N(ca)\vi(ca∈N(vi)), in all non-zero residual errorsIn select
Maximum residul difference
5), calculateAnd set
6), to each ca∈N(vmax)\cmax, calculate
7), to each vb∈N(ca)\vmax, calculate
If 8), all check equations meet or reached the maximum iteration of setting, then terminate decoding, otherwise return
Return step 4.
It can be seen that from above-mentioned arthmetic statement, the LILRBP algorithms and NW RBP and SVNF-RBP algorithm all use check-node
To variable node likelihood ratio absolute residuals localization method, the difference is that the time that LILRBP algorithms difference residual error is produced, only makes
With the residual error of newest generation;And it is indifference pair that the residual error that NW RBP and SVNF-RBP algorithm are produced to different time, which is given,
Treat, all residual errors are made no exception.In view of SVNF-RBP be in current dynamic algorithm it is most representational, therefore LILRBP calculates
Method will be contrasted mainly with SVNF-RBP algorithms.
As shown in figure 1, black circles represent the variable node being updated, black box represents the school being updated
Test node.Dynamic asynchronous information updating strategy step is as follows, first as shown in Fig. 1 (a), selects check-node and disappears to variable node
BreathIt is preferential to update, wherein variable node vjLikelihood ratio symbol will change, then as shown in Fig. 1 (b), to all schools
Test node ca∈N(vj)\ciProduce messageAnd transmitted, finally, Fig. 1 (c), to all check-node ca∈N(vj)\
ci, vb∈N(ci)\vjCalculate residual errorWhich variable node v determined simultaneouslybLikelihood ratio symbol will change.
Fig. 2 lists NW RBP, SVNF-RBP and LILRBP including Flooding BP, LBP in code length 1944,
Code checkThe iterations threshold value I of FER performance maps, wherein LILRBP algorithms when given signal to noise ratio is 1.75dBthrIt is set as
10.It can be seen that LILRBP performance curves in 5 iteration are obvious under SVNF-RBP algorithm curves, and
SVNF-RBP is most representative dynamic decoding algorithm, therefore the main and SVNF-RBP algorithms in analogous diagram below at present
Contrasted.All emulation are carried out under awgn channel.
Fig. 3 gives code length 1944, code checkWhen, respectively 5, under 50 iterationses, SVNF-RBP and LILRBP algorithms
FER performance maps, wherein LILRBP algorithms iterations threshold value IthrIt is set as 10.It can be seen that LILRBP algorithms
In less 5 iteration of iterations, decoding performance has been slightly better than SVNF-RBP algorithms, and performance is substantially better than during high s/n ratio
SVNF-RBP algorithms, in 50 iteration in the case of low signal-to-noise ratio, performance and SVNF-RBP algorithms are basically identical, middle and high
SVNF-RBP algorithms are surmounted during signal to noise ratio, especially in high s/n ratio, performance far wins SVNF-RBP algorithms.
The present invention discloses a kind of local contrary sign dynamic decoding method --- LILRBP (Local based on LDPC code
Inverse and Local Residual BP), this method can not only reduce search and the storage complexity of residual error, also simultaneously
The decoding performance (SVNF-RBP algorithms are in smaller iterations almost without performance boost) in smaller iterations can be lifted,
Decoding performance is also significantly better than SVNF-RBP algorithms during high s/n ratio.
It is ageing that LILRBP algorithms think that residual error has, that is, the residual error that different time is produced, and it acts on also phase not to the utmost
Together.LILRBP algorithms only focus on the residual error of newest generation, and set iterations threshold value, when iterations is less than threshold value, then
In the residual error of newest generation, residual error is screened according to correlated variables node likelihood ratio sign change situation, found out most
Big residual error sets up check-node to the renewal sequence of variable node message;When iterations is higher than threshold value, directly newest
Maximum residul difference is found out in the residual error of generation to set up check-node to the renewal sequence of variable node message.Due to newest generation
Residual error is, by newest information updating generation, therefore always to can ensure that and gone to update next message with newest message, play plus
The convergent effect of speed, and screened again with reference to variable node likelihood ratio sign change situation, further reinforcing accelerates convergence
Effect, improve decoding performance.Meanwhile, maximum residul difference is searched in newest residual error, can be carried out while calculating, therefore be not required to
Residual error is stored, the scope of search reduces, and effectively reduces the storage search complexity of dynamic decoding algorithm.
Above-described embodiment is preferably embodiment, but embodiments of the present invention are not by above-described embodiment of the invention
Limitation, other any Spirit Essences without departing from the present invention and the change made under principle, modification, replacement, combine, simplification,
Equivalent substitute mode is should be, is included within protection scope of the present invention.
Claims (4)
1. a kind of local contrary sign dynamic BP interpretation method based on LDPC code, it is characterised in that iterations threshold value is set, most
In the range of the local residual error newly produced, when iterations is less than threshold value, with reference to the likelihood ratio sign change of correlated variables node
Local residual error is screened again, the update sequence that maximum residul difference determines message is found out in the residual error filtered out;When iteration time
When number is higher than threshold value, the update sequence that the maximum determines message is found out in local residual error.
2. the local contrary sign dynamic BP interpretation method according to claim 1 based on LDPC code, it is characterised in that selected to want
The check-node of renewal is to variable node message(wherein variable node vjLikelihood ratio symbol update after will become
Change), new information firstTo all check-node ca∈N(vj)\ciProduce and transmit messageTo all variable sections
Point vb∈N(ca)\vjCalculate messageLocal residual error, iterations be less than iteration threshold when combine correlated variables node seemingly
Right ratio sign change is screened again to local residual error, is then found out maximum residul difference and is determined next check-node to be updated to change
Measure node messages;When iterations is more than iteration threshold, the maximum is directly found out in local residual error and determines next to update
Check-node to variable node message;Wherein, N (vj) represent and variable node vjConnected all check-node set, N
(vj)\ciRepresent and variable node vjConnected is all except check-node ciOuter check-node, check-node ca∈N(vj)\ci
Represent check-node caTo take all over variable node vjConnected is all except check-node ciOuter check-node;N(ca) represent with
Check-node caConnected all variable nodes, N (ca)\vjRepresent and check-node caConnected is all except variable node vjOuter
Variable node, vb∈N(ca)\vjRepresent variable node vbTo take all over check-node caConnected is all except variable node vjOuter
Variable node.
3. the local contrary sign dynamic BP interpretation method according to claim 2 based on LDPC code, it is characterised in that
Local contrary sign dynamic decoding algorithm (LILRBP):
1) all m are initializedc,v=0;
2) initialize all
3) all r (m are calculatedc,v);
4) in all r (mc,v) in find out the related residual for the v that likelihood ratio symbol can change, and find out wherein5) calculateWithca∈N(vmax)\cmax;
6) calculateca∈N(vmax)\cmax,vb∈N(ca)\vmax;
7) I < IthrWhen, allIn find out the v that likelihood ratio symbol can changebRelated residual, and at it
In find outI > IthrWhen, allIn find out
If 8) all check equations meet or reached the maximum iteration of setting, terminate decoding, otherwise return to step
5);Wherein:mc,vRefer to all check-nodes to the message of variable node;Refer to variable node vnTo all connected verifications
The message of node;Represent variable node viChannel prior probability;r(mc,v) refer to all check-nodes to variable node
Residual error before information updating and after renewal;Represent check-node cjTo variable node viLikelihood ratio update before and more
Residual error after new;N(vi) represent and variable node viConnected all check-node set, N (vi)\cjRepresent and variable node vi
Connected is all except check-node cjOuter check-node, check-node ca∈N(vi)\cjRepresent check-node caTo take all over change
Measure node viConnected is all except check-node cjOuter check-node;Represent from check-node caTo variable node vi's
Message,Represent from variable node viTo check-node cjMessage;N(cj) represent and check-node cjConnected all changes
Measure node, N (cj)\viRepresent and check-node cjConnected is all except variable node viOuter variable node, vb∈N(cj)\vi
Represent variable node vbTo take all over check-node cjConnected is all except variable node viOuter variable node;I represents iteration time
Number, and IthrIt is iterations threshold value.
4. the local contrary sign dynamic BP interpretation method according to claim 3 based on LDPC code, it is characterised in that by seeking
Local maxima residual error is looked for set up a check-node to the renewal sequence of variable node message, likelihood ratio message residual computations public affairs
Formula:r(mk)=| | fk(m)-mk||∞, mk∈m;Wherein m represents to calculate fk(m) related news needed for, mkAnd fk(m) generation respectively
Likelihood ratio before table check-node updates to variable node and after renewal.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710270488.3A CN107241104B (en) | 2017-04-24 | 2017-04-24 | Local different-sign dynamic BP decoding method for LDPC code |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710270488.3A CN107241104B (en) | 2017-04-24 | 2017-04-24 | Local different-sign dynamic BP decoding method for LDPC code |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107241104A true CN107241104A (en) | 2017-10-10 |
CN107241104B CN107241104B (en) | 2020-12-29 |
Family
ID=59984093
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710270488.3A Active CN107241104B (en) | 2017-04-24 | 2017-04-24 | Local different-sign dynamic BP decoding method for LDPC code |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107241104B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110022159A (en) * | 2019-03-12 | 2019-07-16 | 西北工业大学 | A kind of LDPC code decoding algorithm of fast convergence |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7653867B2 (en) * | 2005-03-01 | 2010-01-26 | The Texas A&M University System | Multi-source data encoding, transmission and decoding using Slepian-Wolf codes based on channel code partitioning |
US20100293432A1 (en) * | 2009-05-13 | 2010-11-18 | Yoshikuni Miyata | Error-correction coding method, error-correction decoding method, error-correction coding apparatus, and error-correction decoding apparatus |
-
2017
- 2017-04-24 CN CN201710270488.3A patent/CN107241104B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7653867B2 (en) * | 2005-03-01 | 2010-01-26 | The Texas A&M University System | Multi-source data encoding, transmission and decoding using Slepian-Wolf codes based on channel code partitioning |
US20100293432A1 (en) * | 2009-05-13 | 2010-11-18 | Yoshikuni Miyata | Error-correction coding method, error-correction decoding method, error-correction coding apparatus, and error-correction decoding apparatus |
Non-Patent Citations (1)
Title |
---|
赵明: "改进的基于剩余度置信度传播的LDPC解码算法", 《北京航空航天大学学报》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110022159A (en) * | 2019-03-12 | 2019-07-16 | 西北工业大学 | A kind of LDPC code decoding algorithm of fast convergence |
CN110022159B (en) * | 2019-03-12 | 2021-08-10 | 西北工业大学 | Fast-convergence LDPC decoding algorithm |
Also Published As
Publication number | Publication date |
---|---|
CN107241104B (en) | 2020-12-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104639178B (en) | A kind of dynamic column renewal interpretation method based on LDPC code | |
JP2533219B2 (en) | Viterbi decoding method and apparatus | |
CN106803759A (en) | Polar yards of effective adaptive decoding method based on Gauss construction | |
CN105933010A (en) | Low-complexity polarization code decryption SCL algorithm based on segmented verification assistance | |
JP5301575B2 (en) | Channel coding apparatus and method in digital broadcast communication system using low density parity check code | |
CN104467874B (en) | A kind of LDPC code dynamic dispatching interpretation method based on oscillatory variation node | |
CN111565051B (en) | Self-learning normalized bias minimum sum decoding method for LDPC code | |
CN110535475B (en) | Hierarchical adaptive normalized minimum sum decoding algorithm | |
CN109309502A (en) | The layering LDPC basic matrix of 5G NR standard handles interpretation method | |
CN111277277B (en) | Method and device for reducing decoding delay of polarization code continuous cancellation table decoding algorithm | |
CN103281090A (en) | Mixed modified weighted bit-flipping LDPC decoding algorithm | |
CN113098530A (en) | LDPC code decoding method based on average cluster residual dynamic scheduling selection strategy | |
CN112332864B (en) | Polarization code decoding method and system for self-adaptive ordered mobile pruning list | |
KR20150137430A (en) | Method and apparatus for decoding a non-binary ldpc code in a communication system | |
CN106936445A (en) | A kind of multielement LDPC code coding method of low complex degree near-maximum-likelihood | |
CN110233628A (en) | The adaptive belief propagation list decoding method of polarization code | |
CN106330201A (en) | Variable node reliability dynamic selection strategy-based multi-system LDPC updating method | |
CN105680876A (en) | LDPC code dynamic asynchronous update method based on V2C dynamic selection strategy | |
CN107241104A (en) | One local contrary sign dynamic BP interpretation method for LDPC code | |
CN115208736A (en) | High-order modulation constellation design method suitable for hybrid automatic repeat request system | |
Lu et al. | Deep learning aided SCL decoding of polar codes with shifted-pruning | |
CN103124181A (en) | Turbo code decoding iteration cease method based on cosine similarity | |
CN107404323A (en) | A kind of LDPC code based on staggeredly ranks message transmission improves decoding algorithm | |
CN105406877A (en) | Decoding method of cyclic code with short code length | |
CN107241103A (en) | A kind of local dynamic station BP interpretation methods for LDPC code |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |