CN105515590B - A kind of effective low complex degree serially offsets list polarization code coding method - Google Patents
A kind of effective low complex degree serially offsets list polarization code coding method Download PDFInfo
- Publication number
- CN105515590B CN105515590B CN201510907710.7A CN201510907710A CN105515590B CN 105515590 B CN105515590 B CN 105515590B CN 201510907710 A CN201510907710 A CN 201510907710A CN 105515590 B CN105515590 B CN 105515590B
- Authority
- CN
- China
- Prior art keywords
- decoding
- probability
- data stream
- input
- random
- 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.)
- Active
Links
Landscapes
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
Abstract
The invention discloses a kind of effective low complex degrees based on random binary data stream serially to offset list polarization code decoding algorithm and its decoding framework, algorithm steps are as follows: be changed into corresponding input probability after the vector to be decoded received is scaled algorithm by channel massage；The input probability value calculated is generated into corresponding random binary data stream, data flow is separately input to two basic dish decoding frames；The conditional probability value that data flow is decoded after the calculating by two groups of basic dish decoding frame mixed nodes；The conditional probability random data stream of four paths is calculated by phase inverter；Four conditional probabilities are passed through one respectively and realize final differentiation path probability with door；By the calculating of feedback module, integrated signal is fed back in basic dish-shaped frame, and data flow decodes frame again by basic dish, and obtains conditional probability value；Obtain optimal path decoding result.Present invention reduces the complexities of system, while improving the performance of random data stream decoding.
Description
Technical field
The present invention relates to the efficient coding of coding techniques polarization code, specially a kind of ground complexity string suitable for polarization code
Row offsets list decoding and its decoding framework.
Background technique
It proposes, polar code is the first kind of channel coding, can almost realize the discrete nothing of symmetrical binary system
The capacity (B-DMCs) of memory channel.Since its lower computation complexity is O (NlogN), wherein N is polarization code length；With
And the decoding architecture of fast Fourier transform Fast Fourier Transformation (FFT) form, it is serial to offset decoding
Successive cancellation (SC) algorithm has become one of most effective polarization decoding algorithm.However compared to most
Maximum-likelihood maximum likelihood (ML) decoder, the serial decoding performance for offsetting decoder still have biggish decline.
In order to reduce the suboptimum Path selection bring performance gap by traditional serial counteracting decoder, decoding is serially offset in list
Algorithm (list SC polar decoder) comes into being.After list (L) is added, the chance of more Path selections is brought.
Simulation result shows that the pole list SC decoder can surmount the bit error rate of famous LDPC code.
The serial major defect for offsetting list decoding device is that, with the increase of list L size, the complexity of decoder is linear
Increase.In other words, in order to realize better performance, serial list decoding device of offsetting must be special by greatly hardware cost
It is not when list L is sizable.In addition, existing serial counteracting list is translated for the application scenarios of certain height noises
Code device noise resisting ability is insufficient.To solve the above-mentioned problems, the invention proposes the polarity solutions based on random binary data stream
Code algorithm, i.e. introducing random data stream is as arithmetic core.Have benefited from the advantages of decoding at random, it is serial to offset list decoding device energy
Enough reach good bit error rate performance, reduces complexity and fault-tolerance.However, directly application is based on random data in practice
The serial counteracting list decoding device of stream will lead to the tight of decoding performance due to the continuous loss of random sequence randomness in operation
Decline again.In the present invention, one kind " the double scheme of probability " DPA (Doubling Probability Approach) is proposed, with
Improve decoding performance.The corresponding hardware structure of above-mentioned decoding algorithm is given simultaneously.
Summary of the invention
In view of the above shortcomings of the prior art, the object of the present invention is to provide a kind of having based on random binary data stream
Effect low complex degree serially offsets list polarization code decoding algorithm and its decoding framework, to reduce the complexity of whole system, simultaneously
Improve the performance of random data stream decoding.
To achieve the above object, the invention adopts the following technical scheme:
A kind of effective low complex degree based on random binary data stream serially offsets list polarization code decoding algorithm, including
Following steps:
(1) the vector y to be decoded that will be receivedi=(y1,...,yN) corresponding by being changed into after channel massage scaling algorithm
Input probabilityValue, wherein α is channel zoom factor, and e is natural number constant, and N is polarization code code length；
(2) the corresponding input probability value of the N number of code word calculated is generated the corresponding length of N group by CPU is 1024
Random binary data stream, i.e., each probability value length be 1024 random binary data flow table show；By this N group number
The input terminal of two basic dish decoding frames of decoding framework, the input data phase of two framings are separately input to according to stream step-by-step
Together；
(3) data flow passes through two groups of log2It is obtained after the calculating of N grades of basic dish decoding frame mixed nodesThe item of decoding
Part probability valueWith
(4) the conditional probability random data stream of four paths is calculated by phase inverterWith
(5) four conditional probabilities are passed through one respectively and realize final differentiation path probability with door；
Circulation tire out multiplication:
By obtain four probability valuesCorresponding two optimal paths of two maximum values are chosen by sequence；
Result is passed through " in the double module of probability " and carries out whole probability amplification, and routing diameter is stored by memory；
Two selected paths are correspondingValue feedback, with an input terminal of door, waits four groups of item of next layer into the step
Part probabilityArrival, constantly circulation is tired multiplies；
(6) by the calculating of feedback module, i decoding value in two paths selected is integrated by certain rule
At a signal, feed back in basic dish-shaped frame, data flow decodes frame again by basic dish, and obtainsCode
Conditional probability valueWith
(7) repeat the above steps (4), (5) and (6), until N number of code word has all been translated, to obtain optimal path decoding
As a result
One kind serially offsetting decoding framework based on the list of random binary data stream, it includes two basic dish decoding frames
Frame, two basic dish decoding frames are separately connected list nucleus module；It further include two phase inverters, two phase inverters difference are simultaneously
It is coupled in the connection line of two basic dish decoding frames and list nucleus module；Table nucleus module is connected with feedback module.
The basic dish decoding frame contains N-1 mixed node module, this N-1 node module is divided into
log22 × 1 form of N grade, the framework are connected with one for providing the external feedback module of deployment information.
The mixed node module is due to using binary data stream as processing data, and mixed node module is single herein
Bit arithmetic, the mixed node module include a single input and door, one two input or door, one two input NOT gate, one
A two inputs XOR gate, a JK flip-flop, two two input data selectors, wherein the input terminal of two input XOR gates is defeated
Enter random data stream, output end connects the input terminal of the two or two input data selector；Two input NOT gates input terminals input with
Machine data flow, the input terminal of output the one or two input data selector of termination；The input of one or two input data selector terminates
Random data stream, output end distinguish the input terminal of order input and door and two inputs or door；Single input and door and two inputs or door
Input terminal connect random data stream, the output end of single input and door and two inputs or door connects the input terminal of JK trigger；JK
The output of trigger terminates the two or two input data selector.
The list nucleus module includes 4 and door, and 4 are respectively connected with sorting module with door, and sorting module connects probability
Double module, the double module of probability connect memory memory module, and memory memory module is separately connected 4 and the door.
The beneficial effects of the present invention are:
Invention introduces random binary data streams serially to offset the calculating core of decoding as polarization code list, in this way
Operation by parallel Data Computation Unit originally, be changed to the unit by single-bit and single-bit probabilities data-flow computation.
Largely reduce the complexity of whole system.Pass through " channel massage scalable scheme " Channel Message simultaneously
(Doubling Probability Approach significantly improves random data to Scaling (CMS) with " probability is double " DPA
Flow the performance of decoding.It is specific as follows:
(1) by probabilistic information representated by random binary data stream instead of the rate of specific gravity in traditional SC decoding, ingeniously
Wonderful operation between using bit, largely reduces the hardware complexity of whole system.The complexity of traditional SC decoder is O
(10qN), complexity of the present invention are O (7LN), and wherein q be to quantify length usually to take 10, L to be that list length takes 1 or 2, the N to be herein
Polarize code length；
(2) present invention by " channel massage scalable scheme " Channel Message Scaling (CMS) with " probability turns over
Times " (Doubling Probability Approach significantly improves the performance of random data stream decoding to DPA；
(3) quantization length q is bigger in the present invention, and the length of binary data stream is bigger, and calculated result is more accurate, corresponding
Delay it is longer.And the Modulatory character of random binary data stream length is strong, data flow length accuracy according to needed for system
Allotment, without influencing system architecture.
(4) give detailed effective low complex degree based on random binary data stream serially offsets list to the present invention simultaneously
Polarization code decoding algorithm, while being designed according to the algorithm and having built complete hardware architecture platform.
Detailed description of the invention
Fig. 1 is the performance of BER comparison in the decoding of 64 bit polarization codes under difference α ∈ [0,1] value；
Fig. 2 be when being scaled 0.5 using traditional SC decoding, direct random data stream SC decoding and channel respectively with
The error bit ability comparison diagram of machine data flow SC decoding；
Fig. 3 is list length when being 2, and the decoding path based on random binary data stream selects schematic diagram；
Fig. 4 is that the random data stream after " probability is double " DPA (Doubling Probability Approach) is translated
The bit error rate comparison diagram of code and tradition SC decoding performance；
Fig. 5 be list length be 2 based on the list of random binary data stream serially offset decoding framework；
Fig. 6 is that 8bit polarization code decodes the basic dish decoding frame based on random binary data stream；
Fig. 7 is the framework map of mixed node module；
Fig. 8 is the frame diagram of list nucleus module.
Specific embodiment
The present invention will be further explained with reference to the accompanying drawing.
1, polarization code decoding algorithm is serially offset in the list based on random binary data stream, wherein list length L=1
1) tradition SC decoder algorithms
Consider a polarization code (N, K, A), wherein N indicates the code length of polarization code, and K indicates effective information number in polarization code, A
Indicate effective information bit set.If the vector to be decoded that receiving end receives is yi=(y1,...,yN), the decoding of receiving end
As a result it is expressed asIf uiIt is not effective information bit, we willZero setting.Otherwise decoding bit can
To indicate are as follows:
Wherein,Define transmission probability.Traditional SC decoder passes through two basic meters
It calculates unit and completes decoding calculating, be referred to as f node unit and g node unit.The calculation formula of the two nodes are as follows:
Wherein a and b is two input datas for indicating entry into each node and calculating.
Traditional SC decoder meets FFT dish decoding rule.
2) the SC decoder algorithms based on random binary data stream
Being different from traditional SC decoder uses rate of specific gravity as decoded operating basis is decoded, and present invention introduces random two
Binary data stream carries out decoding calculating.The calculation formula between rate of specific gravity is revised as the public affairs of the calculating on corresponding probability first
Formula then indicates these probability with the form of random binary data stream.
Random binary data flow table shows that mode is as follows: the binary system random data stream for agreeing to length is generated using CPU.Its
The number proportion of middle bit " 1 " is corresponding probability value.Herein due to data flow have randomness, still only know number
According to the corresponding proportion of bit in stream " 1 ", but its specific location is not known.That is the random binary number of a probability is indicated
It is not unique according to stream.Table 1 gives several different situations that probability is expressed as 0.6 so that data flow length takes 10 as an example:
Table 1
Decoding in the present invention calculates based on probability, therefore decodes formula and be rewritten as following form: wherein representing
WhereinIt represents and differentiates decoding resultConditional probability, push over again in this case
The probability calculation expression formula of above-mentioned f node and g node is as follows:
Prf=Pra(1-Prb)+Prb(1-Pra)
Wherein PraWith PrbIndicate the input probability value in cell node calculating, PrfWith PrgIn indicating that cell node calculates
Output probability value.However the bit conditional probability in differentiating can only obtain by the dish-shaped operation framework that the two formula form
But the bit probabilities that cannot be finally decodedThe bit probabilities formula finally decoded
It is as follows:
When list size L is 1, due to not needing to carry out the comparison between each path, so conditional probability hereinIt can be directly as final decoding bit probabilities, without doing tired multiplication.
3) " channel massage scalable scheme " Channel Message Scaling (CMS)
Since in the high region of signal-to-noise ratio, the channel information that binary data stream represents will lead to declining for its random nature
It is weak.So decoding calculating is carried out instead of traditional rate of specific gravity if directlying adopt random binary data stream, although high degree
On save hardware resource, but decoded bit error rate curve performance differs greatly relative to traditional SC decoding.In the present invention
Random data stream is improved using " channel massage scalable scheme " Channel Message Scaling (CMS) to a certain extent
Decoding performance in scheme.
Receiving end vector yiOriginal channel likelihood ratio can be expressed as LR (yi)=4yi/N0, wherein N0It is the unilateral of system
Noise power." channel massage scalable scheme " introduces a channel zoom factor α ∈ [0,1], then new channel likelihood ratio can
To indicate are as follows:
LR′(yi)=α N0LR(yiThe α y of)=4i
The channel likelihood ratio LR ' (y of new received vectori) for generating suitable for the present invention with binary data flow table
Show the most initial input probability value of probability, formula is as follows:
Channel zoom factor is the real number in α ∈ [0,1] section, and the Rational choice of the coefficient has influenced decoding performance
Superiority and inferiority.Fig. 1 gives the performance of BER comparison in the decoding of 64 bit polarization codes under difference α ∈ [0,1] value, wherein choosing
Random data stream length is 1024.It can be seen that gain is maximum when α=0.5, best performance.
Therefore channel zoom factor is used to decode for 0.5 SC based on binary system random data stream in the present invention.Fig. 2
It gives and is scaled respectively using traditional SC decoding, direct random data stream SC decoding and channel in the case of different polarization code lengths
The error bit ability comparison diagram of random data stream SC decoding when 0.5.
2, polarization code decoding algorithm is serially offset in the list based on random binary data stream, wherein list length L=2
1) based on the list SC decoder algorithms of random binary data stream
When list length be greater than 1 when it is necessary to be selected in each layer of 2L path candidate, optimal L paths, herein I
Cannot be again by conditional probabilityDirectly as the foundation of Path selection, but will be according to final bit probabilitiesCarry out preferentially path selection.
Fig. 3 gives when list length is 2, and the decoding path based on random binary data stream selects schematic diagram.Figure
InIndicate the conditional probability in the j-th strip path of i-th layer of bit decision.Per one layer of judgement downwards, need that tired to multiply new condition general
Rate.Two path candidates have been obtained in first layerWithIt is obtained in the second layer
Four path candidates WithIn the final bit probability values of this four paths
In select two optimum probabilities, two optimal paths (as shown in black surround) the most, and abandon two poor paths (such as grey frame institute
Show)；Continue selected path to differentiate to next layer, relatively preferably two are still selected from four path candidates；And so on,
N-th layer is arrived until calculating, optimal path this decoding result the most is selected in four path candidates.
2) " probability is double " DPA (Doubling Probability Approach)
The size for increasing list L, is substantially the selection in order to increase sub-optimal path, and optimal decoding result is avoided not to exist
All in optimal decoding state in each layer of decoding.The decoding performance of L=2 is than L=1 under identical setting for theoretically
Decoding performance is obviously improved, however we do not see expected result in simulations.This is because sentencing calculating path
It is related to when disconnected probability valueTired multiplication.All due to the probability value that here relates to
It is less than 1 number, so by constantly tiring out multiplied arriveIt is a minimum number much smaller than 1.Due toProbability
It is worth too small, is challenged by the probability value accuracy that random binary data flow table is shown.Result in whole error bit ability
And it is not fully up to expectations.
The present invention is promoted using " probability is double " DPA (Doubling Probability Approach)It calculates
In random performance lose.Since Path selection each time is all that probability value in same layer compares, as long as therefore to retain it opposite
Size information does not need absolute size information.When all in same layerWhen value both less than 0.5, the present invention is unified by one
The probability value of layer all carries out double processing, and as newNext layer is brought into continue to calculate.Specific algorithm indicates such as
Under:
Fig. 4 gives the random data after " probability is double " DPA (Doubling Probability Approach)
Stream decoding and the bit error rate comparison diagram of tradition SC decoding performance, wherein random binary data stream length is can in 1024. figures
To find out, DPA algorithm effectively improves the decoding algorithm performance based on random binary data stream.Decoding of the invention is calculated
Method performance boost has been arrived on traditional SC algorithm performance.
Concrete operations of the present invention are as follows:
Fig. 5 is given list length and is serially offset decoding framework based on the list of random binary data stream for 2.It includes
Two basic dish decoding frames, the output of the two frames areWithTwo basic dish decoding frames are separately connected column
Table nucleus module；It further include two phase inverters, two phase inverters are parallel to two basic dish decoding frames and list core respectively
In the connection line of core module；Table nucleus module is connected with feedback module.Realize that binary system is random by the way that two phase inverters are added
Data flow probability takes complementary operation, to obtainWithThe random data of this four groups of expression probability is circulated
Enter list nucleus module, two optimal solutions, path N1 and N2. are preferentially chosen from four paths by feedback module to two roads
Symbol is translated to obtain in diameterCalculating, basic dish decoding frame is fed back to, to provide theFour paths of code value,
And it screens.From husband's above-mentioned steps until N bit polarization code has all been translated.
Fig. 6 gives the decoded basic dish decoding frame based on random binary data stream of 8bit polarization code.Wherein
7 mixed node modules (needing N-1 mixed node module), this 7 module quilts are contained in basic dish framework box
It is divided into 3 grades and (is divided into log2N grades) 2 × 1 forms.The framework also needs an external feedback module to provide deployment information, it will
It has decodedThrough the combination feedback of certain rule into basic dish-shaped framework, to crack next code
ValueThis is the process of a continuous loop iteration.The complexity of the module is O (7N), and tradition SC is decoded under same case
Dish-shaped basic framework complexity is O (10qN), and wherein q is quantization length, and corresponding random binary data stream length is
2q。
Fig. 7 gives the framework map of mixed node module.Due to decode here dish-shaped data flow trend in, f node
It is consistent with the operation framework of g node, therefore the module is designed according to the new probability formula of f node and g node, has merged f node and g is saved
The function of two calculation formula of point.Different probability data stream is exported by damping factor k selection, i.e. when k=0 exports f node
Calculated result exports g node calculated result when k=1.Mixed node module is due to using binary data stream as processing number
According to mixed node module is single-bit operation herein, and the mixed node module includes a single input and door, one two input
Or door, one two input NOT gate, one two input XOR gate, a JK flip-flop, two two input data selectors, wherein
The input terminal of two input XOR gates inputs random data stream, and output end connects the input terminal of the two or two input data selector；Two
The input terminal for inputting NOT gate inputs random data stream, the input terminal of output the one or two input data selector of termination；One or two is defeated
Enter the input termination random data stream of data selector, output end distinguishes the input terminal of order input and door and two inputs or door；
The input terminal of single input and door and two inputs or door connects random data stream, and single input and the output end of door and two inputs or door are equal
Connect the input terminal of JK flip-flop；The output of JK flip-flop terminates the two or two input data selector.
Fig. 8 gives the frame diagram of list nucleus module.List nucleus module includes 4 and door, and 4 are respectively connected with door
Sorting module, sorting module connect the double module of probability, and the double module of probability connects memory memory module, memory storage
Module is separately connected 4 and the door.It is realized by one with doorCirculation is tired to be multiplied
Operation.By obtain four probability valuesCorresponding two optimal paths of two maximum values are chosen by sequence.Result is led to
Enter and carry out whole probability amplification " in the double module of probability ", and routing diameter is stored by memory.
The framework concrete operations process is as follows:
I. the vector y to be decoded that will be receivedi=(y1,...,yN) corresponding by being changed into after channel massage scaling algorithm
Input probabilityValue, wherein α is channel zoom factor；
Ii. the input probability value calculated is generated into the random binary data that corresponding length is 1024 by CPU
This N group data stream step-by-step is separately input to (two groups of input terminal of two basic dish decoding frames of Fig. 6 decoding framework by stream
The input data of frame is identical)；
Iii. data flow passes through two groups of log2It is obtained after the calculating of N grades of basic dish decoding frame mixed nodesDecoding
Conditional probability valueWith
Iv. the conditional probability random data stream of four paths is calculated by phase inverterWith
V. four conditional probabilities are passed through one respectively and realize final differentiation path probability with door
Circulation tire out multiplication.By obtain four probability valuesThe corresponding two optimal roads of two maximum values are chosen by sequence
Diameter.Result is passed through " in the double module of probability " and carries out whole probability amplification, and routing diameter is stored by memory.It will select
Two paths it is correspondingValue feedback, with an input terminal of door, waits four groups of conditional probability of next layer into the stepArrival, constantly circulation is tired multiplies；
Vi. by the calculating of feedback module, i decoding value in two paths selected is integrated by certain rule
At a signal, feed back in basic dish-shaped frame, data flow decodes frame again by basic dish, and obtainsCode
Conditional probability valueWith
Vii. repeat the above steps IV, step V and step VI, until N number of code word has all been translated.To obtain optimal road
Diameter decodes result
It is 10 that table 2, which gives quantization length q, when corresponding random binary data stream length l is 1024, three kinds of schemes
Complexity and the decoded hardware resource consumption of 8bit.(quantization length q is bigger, and the length of binary data stream is bigger, calculated result
More accurate, corresponding delay is longer.And the Modulatory character of data flow length is strong, data flow length can be accurate according to needed for system
Degree allotment, without influencing system architecture.)
Table 2
The above is only a preferred embodiment of the present invention, it should be pointed out that: for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (1)
1. a kind of effective low complex degree based on random binary data stream serially offsets list polarization code coding method, special
Sign is: the following steps are included:
(1) the vector y to be decoded that will be receivedi=(y1,...,yN) by being changed into corresponding input after channel massage scaling algorithm
ProbabilityValue, wherein α is channel zoom factor, and e is natural number constant, and N is polarization code code length；
(2) by the corresponding input probability value of the N number of code word calculated by CPU generate the corresponding length of N group be 1024 with
Machine binary data stream, i.e., the random binary data flow table that each probability value length is 1024 are shown；This N group data stream is pressed
Position is separately input to the input terminal of two basic dish decoding frames of decoding framework, and the input data of two framings is identical；
(3) data flow passes through two groups of log2It is obtained after the calculating of N grades of basic dish decoding frame mixed nodesThe condition of decoding is general
Rate valueWith
(4) the conditional probability random data stream of four paths is calculated by phase inverterWith
(5) four conditional probabilities are passed through one respectively and realize final differentiation path probability with door；
Circulation tire out multiplication:
By obtain four probability valuesCorresponding two optimal paths of two maximum values are chosen by sequence；
Result is passed through " in the double module of probability " and carries out whole probability amplification, and routing diameter is stored by memory；
Two selected paths are correspondingValue feedback, with an input terminal of door, waits four groups of next layer into the step
Conditional probabilityArrival, constantly circulation is tired multiplies；
(6) by the calculating of feedback module, i decoding value in two paths selected is integrated into a signal, is fed back
To in basic dish decoding frame, data flow decodes frame again by basic dish, and obtainsThe conditional probability value of codeWith
(7) repeat the above steps (4), (5) and (6), until N number of code word has all been translated, to obtain optimal path decoding result
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510907710.7A CN105515590B (en) | 2015-12-09 | 2015-12-09 | A kind of effective low complex degree serially offsets list polarization code coding method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510907710.7A CN105515590B (en) | 2015-12-09 | 2015-12-09 | A kind of effective low complex degree serially offsets list polarization code coding method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105515590A CN105515590A (en) | 2016-04-20 |
| CN105515590B true CN105515590B (en) | 2019-01-25 |
Family
ID=55723305
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510907710.7A Active CN105515590B (en) | 2015-12-09 | 2015-12-09 | A kind of effective low complex degree serially offsets list polarization code coding method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105515590B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3497795A1 (en) * | 2016-08-12 | 2019-06-19 | Telefonaktiebolaget LM Ericsson (PUBL) | Iterative decoding of harq-ir transmissions based on soft output polar code decoders |
| CN106656213B (en) * | 2016-12-22 | 2019-10-11 | 东南大学 | The low complex degree polarization code decomposed based on k sections folds the implementation method of hardware architecture |
| CN106877884B (en) * | 2017-02-01 | 2020-04-28 | 东南大学 | Polar code decoding method for reducing decoding path splitting |
| CN108540140B (en) * | 2017-03-01 | 2021-01-22 | 电信科学技术研究院 | Polar code decoding method and device |
| US10756846B2 (en) * | 2017-03-16 | 2020-08-25 | Qualcomm Incorporated | Distributed feedback architecture for polar decoding |
| CN107231158B (en) * | 2017-05-04 | 2019-12-31 | 西南交通大学 | Polarization code iterative receiver, system and polarization code iterative decoding method |
| GB2563473B (en) * | 2017-06-15 | 2019-10-02 | Accelercomm Ltd | Polar coder with logical three-dimensional memory, communication unit, integrated circuit and method therefor |
| CN108063623B (en) * | 2018-01-05 | 2021-09-10 | 重庆邮电大学 | Serial elimination decoding method of Polar code for reducing complexity |
| CN108599776B (en) * | 2018-04-08 | 2021-04-02 | 中山大学 | Improved list serial offset polar code decoding method |
| CN111200439B (en) * | 2018-11-16 | 2022-05-06 | 华为技术有限公司 | Decoding method, device and equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102122966A (en) * | 2011-04-15 | 2011-07-13 | 北京邮电大学 | Channel-polarization-based encoder for staggered structure duplication code, and encoding and decoding methods thereof |
| CN104079382A (en) * | 2014-07-25 | 2014-10-01 | 北京邮电大学 | Polar code decoder and polar code decoding method based on probability calculation |
| CN104124979A (en) * | 2013-04-27 | 2014-10-29 | 华为技术有限公司 | Polar code decoding method and decoding device |
| CN105049061A (en) * | 2015-04-28 | 2015-11-11 | 北京邮电大学 | Advanced calculation-based high-dimensional polarization code decoder and polarization code decoding method |
-
2015
- 2015-12-09 CN CN201510907710.7A patent/CN105515590B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102122966A (en) * | 2011-04-15 | 2011-07-13 | 北京邮电大学 | Channel-polarization-based encoder for staggered structure duplication code, and encoding and decoding methods thereof |
| CN104124979A (en) * | 2013-04-27 | 2014-10-29 | 华为技术有限公司 | Polar code decoding method and decoding device |
| CN104079382A (en) * | 2014-07-25 | 2014-10-01 | 北京邮电大学 | Polar code decoder and polar code decoding method based on probability calculation |
| CN105049061A (en) * | 2015-04-28 | 2015-11-11 | 北京邮电大学 | Advanced calculation-based high-dimensional polarization code decoder and polarization code decoding method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105515590A (en) | 2016-04-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105515590B (en) | A kind of effective low complex degree serially offsets list polarization code coding method | |
| Yuan et al. | Low-latency successive-cancellation list decoders for polar codes with multibit decision | |
| CN105340183B (en) | A kind of interpretation method and device of polar code | |
| CN107026656B (en) | CRC-assisted medium-short code length Polar code effective decoding method based on disturbance | |
| CN105811998B (en) | A kind of polarization code constructing method and polarization code coding/decoding system based on density evolution | |
| Xiong et al. | Symbol-decision successive cancellation list decoder for polar codes | |
| CN105187073B (en) | A kind of the BP interpretation methods and device of polarization code | |
| Tallini et al. | Transmission time analysis for the parallel asynchronous communication scheme | |
| CN111541517B (en) | List polarization code propagation decoding method | |
| Choukroun et al. | Error correction code transformer | |
| CN103929210B (en) | Hard decision decoding method based on genetic algorithm and neural network | |
| CN104124980B (en) | It is adapted to the high speed secret negotiation method of continuous variable quantum key distribution | |
| CN111490796B (en) | Belief propagation decoding method based on dynamic flip bit | |
| CN109450459B (en) | Polarization code FNSC decoder based on deep learning | |
| CN103220002B (en) | A kind of quasi-cyclic LDPC code constructing method eliminating decoder access conflict | |
| CN110138390A (en) | A kind of polarization code SSCL algorithm decoder based on deep learning | |
| CN113300813A (en) | Attention-based combined source channel method for text | |
| CN106656213B (en) | The low complex degree polarization code decomposed based on k sections folds the implementation method of hardware architecture | |
| CN101478313B (en) | Minimum value computing device for LDPC decoder and constructing method thereof | |
| CN111446973B (en) | Polarization code belief propagation decoding method based on multi-flip bit set | |
| Zhang et al. | High‐Throughput Fast‐SSC Polar Decoder for Wireless Communications | |
| Andi et al. | Fast decoding of polar codes using tree structure | |
| CN110166060A (en) | Height is handled up pipeline-type polarization code BP decoder and its implementation | |
| Liang et al. | Efficient stochastic successive cancellation list decoder for polar codes | |
| CN109639290A (en) | The semi-random grouping supercomposed coding of one kind and interpretation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |