CN108494412A - A kind of multiple-factor amendment LDPC code interpretation method and device based on parameter Estimation - Google Patents
A kind of multiple-factor amendment LDPC code interpretation method and device based on parameter Estimation Download PDFInfo
- Publication number
- CN108494412A CN108494412A CN201810342373.5A CN201810342373A CN108494412A CN 108494412 A CN108494412 A CN 108494412A CN 201810342373 A CN201810342373 A CN 201810342373A CN 108494412 A CN108494412 A CN 108494412A
- Authority
- CN
- China
- Prior art keywords
- node
- check
- variable node
- belief messages
- variable
- 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.)
- Pending
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
- H03M13/1108—Hard decision decoding, e.g. bit flipping, modified or weighted bit flipping
Abstract
The invention discloses a kind of, and the multiple-factor based on parameter Estimation corrects LDPC code interpretation method and device, the present invention is during carrying out check-node update processing, finding out minimum value of the variable node to check-node transmission belief messages, after estimating the sub-minimum that variable node transmits belief messages to check-node, different modifying factors are added before minimum value and sub-minimum to be modified the belief messages of transmission, to efficiently solve in soft-decision LDPC code decoding algorithm, the operation that minimum value looks for sub-minimum again should be found in every a line check-node renewal process of check matrix, the problem of largely effecting on decoding rate and implementation complexity.
Description
Technical field
The present invention relates to fields of communication technology more particularly to a kind of multiple-factor based on parameter Estimation to correct LDPC code decoding
Method and device.
Background technology
The meaning of research LDPC code is not only only that them close to the performance of shannon limit, even if being more it using a series of
The Low Complexity Decoding Algorithm of suboptimum can also obtain good decoding performance.Two major classes are broadly divided into for decoding, it is a kind of
It is the Soft decision decoding algorithm based on belief propagation.Another kind of is the Hard decision decoding algorithm based on bit reversal.
For Hard decision decoding algorithm, LDPC code presenter Gallager initially proposes bit reversal calculation in paper
Method.Later, constantly there is scholar and study bit flipping algorithm with some spreading parameters, error-correcting performance is made to optimize significantly, corrected
Mistake it is more.Although there is numerous scholars to carry out linguistic term to hard decision algorithm, its performance is eventually not as good as soft-decision is translated
Code algorithm.
For Soft decision decoding algorithm, LDPC code presenter Gallager proposes a kind of confidence level in initial paper
The algorithm of propagation.In LDPC code decoding algorithm, BP decoding algorithms are closest to the decoding algorithm of shannon limit, but due to decoding
A large amount of tanh (x) functions, tanh involved in process-1(x) function and multiplying so that computation complexity is very high.It connects down
Carry out a series of scholar to start to simplify BP algorithm.SPA decoding algorithms make the carry out logarithmetics processing of BP decoding algorithms
Largely tired multiplication becomes accumulating operation during decoding, reduces complexity.MS decoding algorithms are SPA decoding algorithms
The tired multiplication of middle tanh (x) function is simplified to minimize the operation of sub-minimum, under the premise of having lost fraction performance,
Greatly reduce complexity.For error caused by MS decoding algorithms approximate calculation in the process, NMS, OMS etc. are a series of for mending
The decoding algorithm for repaying error during approximate calculation is proposed out.
Currently, whether MS algorithms or NMS algorithms and OMS algorithms, are required in the check-node update of every a line
Find minimum value and sub-minimum.In the case where not considering symbol, sub-minimum is put at the position of minimum value, other positions are put
Upper minimum value.For hardware realization, in majority of comforming in seldom clock, minimum value and sub-minimum are found out, and record
The position of minimum value, such operation need complicated compounding-logic circuit.Usually in the iterative decoding of LDPC code, each time
The row update number that iteration needs is very more, is equal to the number of check-node.Therefore, this update per a line should find minimum
Value looks for the operation of sub-minimum that can largely effect on the rate of decoding again.
Invention content
In view of above-mentioned analysis, the present invention is intended to provide a kind of multiple-factor based on parameter Estimation corrects LDPC code decoding side
Method and device, fully or at least partially to solve the above problems.
To solve the above problems, the present invention is mainly achieved through the following technical solutions：
The present invention provides a kind of, and the multiple-factor based on parameter Estimation corrects LDPC code interpretation method, which is characterized in that packet
It includes：Step 1: by the channel information of reception, the belief messages that all variable nodes are transmitted to check-node are initialized；Step
Rapid two, pass through the revised minimum value of modifying factor in the belief messages transmitted by variable node to check-node, and
The sub-minimum of belief messages is obtained by the revised estimation of modifying factor, the confidence transmitted to check-node to variable node
It spends message and carries out approximate calculation, complete the update processing of check-node；Step 3: variable node is transmitted by check-node
Belief messages and the channel massage of reception calculate variable node to the newer belief messages of check-node, complete variable section
The update processing of point；Step 4: the channel massage of the belief messages and reception transmitted to variable node by check-node, meter
Calculate the belief messages of variable node；Step 5: the belief messages to variable node carry out hard decision, one and code length are obtained
Isometric sequence, if the sequence meets the parity check equation of LDPC code or reaches maximum iteration, output should
Sequence after LDPC code decoding, otherwise return to step two, until obtained sequence meets the parity check equation of LDPC code or reaches
Until maximum iteration.
Further, the belief messages that all variable nodes of initialization are transmitted to check-node, specifically include：It is logical
Later probability P is testediIt calculates, the belief messages that all variable nodes are transmitted to check-node is initialized and stored with its result
Lvc, to some information bit i, posterior probabilityWherein, Pr { zi=b | yi, b ∈ { 0,1 } are given
Channel exports yiUnder the conditions of, code word bits z after LDPC codingsiPosterior probability equal to b, when the mean value of channel is 0, variance σ2
Awgn channel, then
Further, modifying factor amendment is passed through in the belief messages transmitted by variable node to check-node
Minimum value afterwards, and by the revised sub-minimum for estimating to obtain of modifying factor, check-node to variable node is transmitted
Belief messages carry out approximate evaluation, complete check-node update processing, specifically include：
To each check-node c according to formula：
Transmit message Rcv
Variable node v is given, wherein：Normalize modifying factor
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvTo receive
Soft Inform ation, σ2For the variance of interchannel noise.
Further, modifying factor amendment is passed through in the belief messages transmitted by variable node to check-node
Minimum value afterwards, and by the revised sub-minimum for estimating to obtain of modifying factor, check-node to variable node is transmitted
Belief messages carry out approximate evaluation, complete check-node update processing, specifically include：
To each check-node c according to formula：
Transmit message Rcv
Variable node v is given, wherein：
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvTo receive
Soft Inform ation, σ2For the variance of interchannel noise.
Further, modifying factor amendment is passed through in the belief messages transmitted by variable node to check-node
Minimum value afterwards, and by the revised sub-minimum for estimating to obtain of modifying factor, check-node to variable node is transmitted
Belief messages carry out approximate evaluation, complete check-node update processing, specifically include：To each check-node c according to
Formula：
It calculates each
Check-node c passes to the belief messages of variable node v, wherein：
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvTo receive
Soft Inform ation, σ2For the variance of interchannel noise；
Further, the channel massage of the belief messages and reception that are transmitted to variable node by check-node calculates
The belief messages of variable node, specifically include：
According toCalculate the belief messages of variable node.
Further, hard decision is carried out to the belief messages of variable node, obtains a sequence isometric with code length, such as
Sequence described in fruit meets the parity check equation of LDPC code or reaches maximum iteration, then exports LDPC code decoding postorder
It arranges, otherwise return to step two, until obtained sequence meets the parity check equation of LDPC code or reaches maximum iteration
Until, it specifically includes：
Hard decision is carried out to the belief messages of variable node, a sequence isometric with code length is obtained, if the sequence
Row meet the parity check equation of LDPC code or reach maximum iteration, then export sequence after LDPC code decoding, otherwise
The sub-minimum that the minimum value of the belief messages transmitted to variable node to check-node, estimation obtain utilizes different modifying factors
After amendment, the belief messages transmitted to variable node using this value approximate calculation check-node complete the update of check-node
Processing, the channel massage of the belief messages and reception that are transmitted using check-node to variable node calculate variable node to school
The belief messages for testing node transmission complete the update processing of variable node, calculate the belief messages of variable node, hard decision
Obtain a sequence isometric with code length afterwards, so recycle, until obtained sequence meet LDPC code parity check equation or
Until person reaches maximum iteration.
Further, hard decision is carried out to the belief messages of variable node, obtains a sequence isometric with code length, such as
Sequence described in fruit meets the parity check equation of LDPC code or reaches maximum iteration, then exports LDPC code decoding postorder
Row, the sub-minimum that the minimum value of the belief messages otherwise transmitted to variable node to check-node, estimation obtain utilize difference
After modifying factor is corrected, the belief messages transmitted to variable node using this value approximate calculation check-node complete verification section
The update processing of point, the channel massage of the belief messages and reception that are transmitted using check-node to variable node calculate variable
The belief messages that node is transmitted to check-node, complete the update processing of variable node, and the confidence level for calculating variable node disappears
It ceases, a sequence isometric with code length is obtained after hard decision, so recycle, until obtained sequence meets the odd even school of LDPC code
Proved recipe journey or until reaching maximum iteration, specifically includes：
If Lv＜ 0, the then code word bits for decoding outOtherwiseIfOr reach repeatedly
Generation number l=Itermax, then iteration stopping, otherwise iterations l=l+1, the confidence that variable node to check-node is transmitted
After the sub-minimum that the minimum value of degree message, estimation obtain is corrected using different modifying factors, section is verified using this value approximate calculation
The belief messages that point is transmitted to variable node are completed the update processing of check-node, are passed using check-node to variable node
The channel massage of the belief messages and reception passed calculates the belief messages that variable node is transmitted to check-node, completes to become
The update processing for measuring node, calculates the belief messages of variable node, a sequence isometric with code length is obtained after hard decision, such as
This cycle, until obtained sequence meets the parity check equation of LDPC code or reaches maximum iteration.
On the other hand, the present invention also provides a kind of, and the multiple-factor based on parameter Estimation corrects LDPC code code translator, the dress
Set including：Initialization unit initializes all variable nodes and is set to what check-node transmitted for the channel information by receiving
Reliability message；Processing unit is updated, modifying factor is passed through in the belief messages for being transmitted to check-node by variable node
The revised minimum value of son, and the sub-minimum of belief messages is obtained by the revised estimation of modifying factor, verification is saved
Point carries out approximate evaluation estimation to the belief messages that variable node transmits, and completes the update processing of check-node；Pass through verification
The channel massage for the belief messages and reception that node transmits variable node calculates variable node and is set to check-node is newer
Reliability message completes the update processing of variable node；The belief messages that variable node is transmitted by check-node and reception
Channel massage, calculate the belief messages of variable node；Decision process unit carries out the belief messages of variable node hard
Judgement, obtains a sequence isometric with code length, if the sequence meets the parity check equation of LDPC code or reaches most
Big iterations then export sequence after LDPC code decoding, and otherwise iterations add one, and restart to calculate, until obtaining
Sequence meet the parity check equation of LDPC code or reach maximum iteration until.
Further, the decision process unit is additionally operable to, to each check-node c according to formula：
Transmit message Rcv
Variable node v is given, wherein：Normalize modifying factor
Alternatively,
To each check-node c according to formula：
Transmit message Rcv
Variable node v is given, wherein：
Alternatively,
To each check-node c according to formula
Wherein：
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvTo receive
Soft Inform ation, σ2For the variance of interchannel noise；According toThe belief messages of variable node are calculated, if
Lv＜ 0, the then code word bits for decoding outOtherwiseIfOr reach iterations l=
Itermax, then iteration stopping, otherwise l=l+1 are then back to the first step, and by revised code word bitsOutput.
The present invention has the beneficial effect that：
The present invention transmits confidence during carrying out check-node update processing, finding out variable node to check-node
The minimum value for spending message, after estimating the sub-minimum that variable node transmits belief messages to check-node, minimum value with it is secondary
Different modifying factors is added before small value to be modified the belief messages of transmission, to efficiently solve soft-decision LDPC code
In decoding algorithm, the fortune that minimum value looks for sub-minimum again should be found in every a line check-node renewal process of check matrix
The problem of calculating, largely effecting on decoding rate and implementation complexity.
Other features and advantages of the present invention will illustrate in the following description, and partial become from specification
It is clear that understand through the implementation of the invention.The purpose of the present invention and other advantages can by the specification write,
Specifically noted structure is realized and is obtained in claims and attached drawing.
Description of the drawings
Fig. 1 is that the flow that a kind of multiple-factor based on parameter Estimation of the embodiment of the present invention corrects LDPC code interpretation method is shown
It is intended to；
Fig. 2 a) it is that the embodiment of the present invention passes through (8176,7154) LDPC code, BPSK modulation decodes after awgn channel transmission
The equal iteration of algorithm 10 times, when SNR is 3.6dB, sub-minimum is corrected with alpha1, and minimum value is three-dimensional with the modified bit error rates of alpha2
Figure, wherein x, y, z coordinate indicates modifying factor alpha2, alpha1 and the bit error rate respectively；
Fig. 2 b) be Fig. 2 a of the embodiment of the present invention) contour map form, x, y indicate respectively modifying factor alpha2,
Alpha1, different contours represent the bit error rate；
Fig. 3 a) it is that the embodiment of the present invention is encoded by (8176,7154) LDPC code, BPSK modulation, after awgn channel transmission,
The equal iteration of decoding algorithm 10 times, when SNR is 3.8dB, sub-minimum is corrected with alpha1, the minimum value modified bit error rates of alpha2
Graphics, wherein x, y, z coordinate indicate modifying factor alpha2, alpha1 and the bit error rate respectively；
Fig. 3 b) be Fig. 3 a of the embodiment of the present invention) contour map form, x, y indicate respectively modifying factor alpha2,
Alpha1, different contours represent the bit error rate；
Fig. 4 is that the embodiment of the present invention is encoded by (8176,7154) LDPC code, and BPSK modulation carries after awgn channel transmission
The ber curve of the decoding algorithm (IEMS) and existing EMS decoding algorithms based on parameter Estimation gone out；
Fig. 5 is that the embodiment of the present invention is encoded by (8176,7154) LDPC code, and BPSK modulation is translated after awgn channel transmission
Code maximum iteration is set as 10 times, and decoding algorithm (IEMS), existing EMS decoding algorithms and MS based on parameter Estimation are calculated
The bit error rate performance comparison diagram of method；
Multiple-factor amendment of Fig. 6 embodiment of the present invention based on parameter Estimation decodes (IENMS) algorithm (its sub-minimum modifying factor
Sub- alpha1=0.7, minimum value modifying factor alpha2=0.65) with multiple-factor is added in the decoding algorithm of existing parameter Estimation
Correct the comparison diagram of decoding (ENMS) algorithm；
Multiple-factor amendment of Fig. 7 embodiment of the present invention based on parameter Estimation decodes (IENMS) algorithm (its sub-minimum modifying factor
Sub- alpha1=0.7, minimum value modifying factor alpha2=0.65), multiple-factor is added in the decoding algorithm of existing parameter Estimation
Modified decoding (ENMS) algorithm, multiple-factor correct minimum and decoding (INMS) algorithm and minimum and decoding (MS) algorithm comparison
Figure；
Fig. 8 is that the structure that a kind of multiple-factor based on parameter Estimation of the embodiment of the present invention corrects LDPC code code translator is shown
It is intended to.
Specific implementation mode
Specifically describing the preferred embodiment of the present invention below in conjunction with the accompanying drawings, wherein attached drawing constitutes the application part, and
It is used to illustrate the principle of the present invention together with embodiments of the present invention.For purpose of clarity and simplification, when it may make the present invention
Theme it is smudgy when, illustrating in detail for known function and structure in device described herein will be omitted.
The multiple-factor based on parameter Estimation that an embodiment of the present invention provides a kind of correcting LDPC code interpretation method, and the present invention is real
It applies example to overcome in existing soft-decision LDPC code decoding algorithm, all minimum value should be found again in the update processing of check-node
The problem of looking for the operation of sub-minimum to largely effect on decoding rate and implementation complexity makes decoding while reducing operation, maximum
Limit ensures the correctness of LDPC code decoding.Below in conjunction with attached drawing and several embodiments, the present invention is carried out further detailed
Explanation.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, the present invention is not limited.
The multiple-factor that an embodiment of the present invention provides a kind of based on parameter Estimation corrects LDPC code interpretation method, referring to Fig. 1,
This method includes：
Step 1: by the channel information of reception, initializes the confidence level that all variable nodes are transmitted to check-node and disappear
Breath；
Step 2: revised most by modifying factor in the belief messages transmitted by variable node to check-node
Small value, and the sub-minimum of belief messages is obtained by the revised estimation of modifying factor, to check-node to variable node
The belief messages of transmission carry out approximate calculation, complete the update processing of check-node；
Step 3: the channel massage of the belief messages and reception transmitted to variable node by check-node, calculates and becomes
Node is measured to the newer belief messages of check-node, completes the update processing of variable node；
Step 4: the channel massage of the belief messages and reception transmitted to variable node by check-node, calculates and becomes
Measure the belief messages of node；
Step 5: the belief messages to variable node carry out hard decision, a sequence isometric with code length is obtained, if
The sequence meets the parity check equation of LDPC code or reaches maximum iteration, then exports LDPC code decoding postorder
It arranges, otherwise return to step two, until obtained sequence meets the parity check equation of LDPC code or reaches maximum iteration
Until.
That is, the embodiment of the present invention arrives during carrying out check-node update processing finding out variable node
Check-node transmits the minimum value of belief messages, estimates the sub-minimum that variable node transmits belief messages to check-node
Afterwards, different modifying factors are added before minimum value and sub-minimum to be modified the belief messages of transmission, to effectively solve
It has determined in soft-decision LDPC code decoding algorithm, minimum value should be found again in every a line check-node renewal process of check matrix
The operation for looking for sub-minimum, the problem of largely effecting on decoding rate and implementation complexity.
Further, when it is implemented, all variable nodes of initialization described in the embodiment of the present invention are transmitted to check-node
Belief messages, specifically include：Pass through posterior probability PiIt calculates, is initialized with its result and store all variable nodes and arrived
The belief messages L that check-node transmitsvc, to some information bit i, posterior probabilityWherein, Pr
{zi=b | yi, b ∈ { 0,1 } are that given channel exports yiUnder the conditions of, code word bits z after LDPC codingsiPosterior probability equal to b,
When the mean value of channel is 0, variance σ2Awgn channel, then
Further, when it is implemented, the confidence transmitted by variable node to check-node described in the embodiment of the present invention
It spends and passes through the revised minimum value of modifying factor in message, and the sub-minimum obtained by the revised estimation of modifying factor,
Approximate evaluation is carried out to the belief messages that check-node to variable node transmits, completes the update processing of check-node, specifically
Including：
To each check-node c according to formula：
Transmit message Rcv
Variable node v is given, wherein：Normalize modifying factor
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvTo receive
Soft Inform ation, σ2For the variance of interchannel noise.
Further, described to be set by what variable node to check-node transmitted when it is implemented, in the embodiment of the present invention
Pass through the revised minimum value of modifying factor in reliability message, and by modifying factor it is revised estimate to obtain it is time small
Value carries out approximate evaluation to the belief messages that check-node to variable node transmits, and completes the update processing of check-node, tool
Body includes：
To each check-node c according to formula：
Transmit message Rcv
Variable node v is given, wherein：
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvTo receive
Soft Inform ation, σ2For the variance of interchannel noise.
Further, described to be set by what variable node to check-node transmitted when it is implemented, in the embodiment of the present invention
Pass through the revised minimum value of modifying factor in reliability message, and by modifying factor it is revised estimate to obtain it is time small
Value carries out approximate evaluation to the belief messages that check-node to variable node transmits, and completes the update processing of check-node, tool
Body includes：
To each check-node c according to formula
Calculate each school
The belief messages that node c passes to variable node v are tested, wherein：
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvTo receive
Soft Inform ation, σ2For the variance of interchannel noise；
The channel of the belief messages and reception that are transmitted to variable node by check-node described in the embodiment of the present invention disappears
Breath, calculates the belief messages of variable node, specifically includes：
According toCalculate the belief messages of variable node.
Described in the embodiment of the present invention to the belief messages of variable node carry out hard decision, obtain one it is isometric with code length
Sequence exports the LDPC code if the sequence meets the parity check equation of LDPC code or reaches maximum iteration
Sequence after decoding, otherwise return to step two, until obtained sequence meets the parity check equation of LDPC code or reaches maximum
Until iterations, specifically include：
Hard decision is carried out to the belief messages of variable node, a sequence isometric with code length is obtained, if the sequence
Row meet the parity check equation of LDPC code or reach maximum iteration, then export sequence after LDPC code decoding, otherwise
The sub-minimum that the minimum value of the belief messages transmitted to variable node to check-node, estimation obtain utilizes different modifying factors
After amendment, the belief messages transmitted to variable node using this value approximate calculation check-node complete the update of check-node
Processing, the channel massage of the belief messages and reception that are transmitted using check-node to variable node calculate variable node to school
The belief messages for testing node transmission complete the update processing of variable node, calculate the belief messages of variable node, hard decision
Obtain a sequence isometric with code length afterwards, so recycle, until obtained sequence meet LDPC code parity check equation or
Until person reaches maximum iteration.
Further, in the embodiment of the present invention, hard decision is carried out to the belief messages of variable node, obtains one and code
Long isometric sequence exports if the sequence meets the parity check equation of LDPC code or reaches maximum iteration
Sequence after LDPC code decoding, the minimum value of the belief messages otherwise transmitted to variable node to check-node, estimation obtain
Sub-minimum corrected using different modifying factors after, the confidence level transmitted to variable node using this value approximate calculation check-node
Message completes the update processing of check-node, the letter of the belief messages and reception that are transmitted using check-node to variable node
Road message calculates the belief messages that variable node is transmitted to check-node, completes the update processing of variable node, calculates variable
The belief messages of node obtain a sequence isometric with code length after hard decision, so recycle, until obtained sequence meets
The parity check equation of LDPC code or until reaching maximum iteration, specifically includes：
If Lv＜ 0, the then code word bits for decoding outOtherwiseIfOr reach repeatedly
Generation number l=Itermax, then iteration stopping, otherwise iterations l=l+1, the confidence that variable node to check-node is transmitted
After the sub-minimum that the minimum value of degree message, estimation obtain is corrected using different modifying factors, section is verified using this value approximate calculation
The belief messages that point is transmitted to variable node are completed the update processing of check-node, are passed using check-node to variable node
The channel massage of the belief messages and reception passed calculates the belief messages that variable node is transmitted to check-node, completes to become
The update processing for measuring node, calculates the belief messages of variable node, a sequence isometric with code length is obtained after hard decision, such as
This cycle, until obtained sequence meets the parity check equation of LDPC code or reaches maximum iteration.
Assuming that parity check matrix H is M × N-dimensional, and Hc,vIndicate the c rows in H, v row.N (c)={ v:Hc,v=1 }
Indicate all variable node v, M (v)={ c being connected with check-node c:Hc,v=1 } it indicates all to be connected with variable node v
Check-node c.N (c) v indicate set N (c) remove variable node v, M (v) c indicate set M (v) removing check-nodes c.
Present embodiment is illustrated in conjunction with the flow chart of Fig. 1, the multiple-factor proposed by the present invention based on parameter Estimation is repaiied
Positive interpretation method, is realized by following steps：
Step1：Initialization
Receiving channel information initializes decoder, and maximum iteration Itermax is arranged, and initially repeatedly
Generation number is set as l=1
Wherein, Pr { zv=b | rv, b ∈ { 0,1 } are that given channel exports rvUnder the conditions of, code word bits z after LDPC codingsv
Posterior probability equal to b.If the mean value of channel is 0, variance σ2Awgn channel, then
Step2：Iterative process
RcvRepresent the belief messages that check-node c is transmitted to variable node v, LvcVariable node v is represented to check-node
The belief messages that c is transmitted.The two message transmissions can be calculated according to (2) and (3).
Step2.1：The update of check-node is handled
Each check-node c transmits message R according to (2)cvGive variable node v
Wherein：
Step2.2：The update of variable node is handled
Each variable node v transmits L according to (3)vcGive check-node c:
Wherein：rvIndicate the Soft Inform ation received, σ2Indicate the variance of interchannel noise.
Step3：Calculate the belief messages of variable node
Step4：Hard decision and iteration stopping condition criterion
If Lv＜ 0, the then code word bits for decoding outOtherwise
IfOr reach iterations l=Itermax, then iteration stopping.
Otherwise l=l+1 is then back to the first step.
Fig. 2 a) it is the BPSK modulation by (8176,7154) LDPC code, after awgn channel transmission, the equal iteration of decoding algorithm 10
Secondary, when SNR is 3.6dB, sub-minimum is corrected with alpha1, the minimum value modified bit error rate graphics of alpha2, wherein x, y, z
Coordinate indicates modifying factor alpha2, alpha1 and error rate BER respectively, is as can be seen from the figure combined in different modifying factors
Under, ber curve has notable difference.
Fig. 2 b) be Fig. 2 a) contour map form, x, y indicate modifying factor alpha2, alpha1, contour respectively
Color represent error rate BER.Fig. 2 b) combine Fig. 2 a), it can clearly be seen that in modifying factor alpha1 in [0.65~0.8],
Combinations of the alpha2 between [0.5~0.7], and sub-minimum modifying factor alpha2 is relative to minimum value modifying factor alpha1
In the case of smaller, the bit error rate is intended to lower.
Fig. 3 a) it is to be encoded by (8176,7154) LDPC code, BPSK modulation, after awgn channel transmission, decoding algorithm changes
In generation 10 times, when SNR is 3.8dB, sub-minimum is corrected with alpha1, the minimum value modified bit error rate graphics of alpha2, wherein
X, y, z coordinate indicates modifying factor alpha2, alpha1 and error rate BER respectively, as can be seen from the figure in different modifying factors
Under sub-portfolio, ber curve has notable difference.
Fig. 3 b) be Fig. 3 a) contour map form, x, y indicate modifying factor alpha2, alpha1, contour respectively
Color represent error rate BER.Fig. 3 b) combine Fig. 3 a), it can clearly be seen that in modifying factor alpha1 in [0.65~0.8],
Combinations of the alpha2 between [0.5~0.7], and sub-minimum modifying factor alpha2 is relative to minimum value modifying factor alpha1
In the case of smaller, the bit error rate is intended to lower.
Fig. 4 be by (8176,7154) LDPC code encode, BPSK modulation, awgn channel transmission after, proposition based on parameter
The ber curve of the decoding algorithm (IEMS) and existing EMS decoding algorithms of estimation.It can be seen that IEMS proposed by the present invention
Algorithm has the gain of about 0.7dB, either original EMS algorithms or IEMS algorithms compared with EMS algorithms, iteration 10 times and iteration 20
Secondary performance difference is simultaneously little, therefore next performance compares iterations and is disposed as 10 times.
Fig. 5 is encoded by (8176,7154) LDPC code, BPSK modulation, after awgn channel transmission, decoding greatest iteration time
Number is set as 10 times, and decoding algorithm (IEMS), the error code of existing EMS decoding algorithms and MS algorithms based on parameter Estimation are forthright
It can comparison diagram.As can be seen that IEMS algorithms only have the performance loss of about 0.7dB compared with MS algorithms, and EMS algorithms have about compared with MS algorithms
The performance loss of 1.4dB.
Fig. 6 simulated conditions are identical as Fig. 5, it indicates that the multiple-factor amendment based on parameter Estimation decodes (IENMS) algorithm (its
Sub-minimum modifying factor alpha1=0.7, minimum value modifying factor alpha2=0.65) with the decoding algorithm of existing parameter Estimation
The middle comparison diagram that multiple-factor amendment decoding (ENMS) algorithm is added.It is clear that IENMS algorithms proposed by the present invention compared with
ENMS algorithms have the gain of about 0.5dB.
Fig. 7 simulated conditions are identical as Fig. 6, it indicates that the multiple-factor amendment based on parameter Estimation decodes (IENMS) algorithm (its
Sub-minimum modifying factor alpha1=0.7, minimum value modifying factor alpha2=0.65), the decoding algorithm of existing parameter Estimation
Modified decoding (ENMS) algorithm of middle addition multiple-factor, multiple-factor correct minimum and decoding (INMS) algorithm and minimum and decoding
(MS) comparison diagram of algorithm.It can be seen that IENMS algorithms compared with MS algorithms only have about 0.2dB performance loss, and ENMS algorithms compared with
MS algorithms have the performance loss of about 0.8dB.IENMS algorithms have the performance loss of 0.3dB compared with INMS algorithms, but can reduce
Hardware realization complexity, this embodies the multiple-factor proposed by the present invention based on parameter Estimation and corrects the low hard of interpretation method realization
The compromise of part complexity and decoding performance.
Table 1 is complexity of the various Soft decision decoding algorithms in the case where log-domain is estimated.Wherein：N indicates that code length, M indicate school
Test length, the d of positionrIndicate row weight, dcIndicate row weight.Since comparison operation is similar with (subtracting) method operand is added, in table 1, compare
Operation is all counted as add operation.It is proposed by the present invention to utilize the original standard of the method for noise variance and minimum estimation sub-minimum ratio
The method for really calculating sub-minimum reducesSub-addition operation, but have different degrees of performance loss.Its
In：Performance loss of the IEMS algorithms compared with MS algorithms is about 0.8dB；Performance loss of the IENMS algorithms compared with NMS algorithms is about 0.4dB；
Performance loss of the IENMS algorithms compared with NMS algorithms is about 0.3dB.From reductionSub-addition can see, row
It is again bigger, in the case that code check is lower, code length is longer, the complexity that this algorithm reduces is more.Present invention emulation use-case is adopted
For (8176,7154) code, in the case of iteration 10 times, the addition number of reduction is 40880 times.
On the whole, the method for the embodiment of the present invention is considered by the compromise between decoding complexity and performance.
For the realization of high-speed coding, reduction is not only complexity, be limited to that minimum value and sub-minimum find when
Clock can also be improved accordingly, be beneficial to the realization of high-speed coding.Therefore, for needing high-speed coding, but to decoding performance requirement
It can extensive use under less high occasion.
Corresponding with Fig. 1, the embodiment of the present invention additionally provides a kind of multiple-factor amendment LDPC code based on parameter Estimation
Code translator, referring to Fig. 8, including：
Initialization unit initializes what all variable nodes were transmitted to check-node for the channel information by receiving
Belief messages；
Processing unit is updated, modifying factor is passed through in the belief messages for being transmitted to check-node by variable node
Revised minimum value, and the sub-minimum of belief messages is obtained by the revised estimation of modifying factor, to check-node
The belief messages transmitted to variable node carry out approximate evaluation estimation, complete the update processing of check-node；It is saved by verifying
The channel massage for the belief messages and reception that point transmits variable node calculates variable node to the newer confidence of check-node
Message is spent, the update processing of variable node is completed；Variable node is transmitted by check-node belief messages and reception
Channel massage calculates the belief messages of variable node；
Decision process unit carries out hard decision to the belief messages of variable node, obtains a sequence isometric with code length
Row, if the sequence meets the parity check equation of LDPC code or reaches maximum iteration, export the LDPC code and translate
Sequence after code, otherwise iterations add one, and restart to calculate, until obtained sequence meets the even-odd check side of LDPC code
Journey or until reaching maximum iteration.
That is, the present invention during carrying out check-node update processing, is saved finding out variable node to verification
Point transmits the minimum value of belief messages, after estimating the sub-minimum that variable node transmits belief messages to check-node,
Different modifying factors are added before minimum value and sub-minimum to be modified the belief messages of transmission, it is soft to efficiently solve
Adjudicate in LDPC code decoding algorithm, every a line check-node renewal process of check matrix should find minimum value look for again it is secondary
The operation of small value, the problem of largely effecting on decoding rate and implementation complexity.
Further, decision process unit described in the embodiment of the present invention be by each check-node c according to formula：Transmit message RcvGive variable section
Point v, wherein：Normalize modifying factor
Alternatively, to each check-node c according to formula：
Transmit message Rcv
Variable node v is given, wherein：
Alternatively, to each check-node c according to formula
Wherein：
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvTo receive
Soft Inform ation, σ2For the variance of interchannel noise；According toThe belief messages of variable node are calculated, if
Lv＜ 0, the then code word bits for decoding outOtherwiseIfOr reach iterations l=
Itermax, then iteration stopping, otherwise l=l+1 are then back to the first step, and by revised code word bitsOutput.
On the whole, the embodiment of the present invention is considered by the compromise between decoding complexity and performance.Simultaneously for height
For the realization of speed decoding, reduction is not only complexity, and being limited to clock that minimum value is found with sub-minimum can also phase
It should improve, be beneficial to the realization of high-speed coding.Therefore, for needing high-speed coding, but it is less high to decoding performance requirement
It can extensive use under occasion.
The related content of the embodiment of the present invention can refer to embodiment of the method part and be understood, not be described in detail herein.
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto,
Any one skilled in the art in the technical scope disclosed by the present invention, the change or replacement that can be readily occurred in,
It should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with the protection model of claims
Subject to enclosing.
Claims (10)
1. a kind of multiple-factor based on parameter Estimation corrects LDPC code interpretation method, which is characterized in that including：
Step 1: by the channel information of reception, the belief messages that all variable nodes are transmitted to check-node are initialized；
Step 2: passing through the revised minimum of modifying factor in the belief messages transmitted by variable node to check-node
Value, and the sub-minimum of belief messages is obtained by the revised estimation of modifying factor, check-node to variable node is passed
The belief messages passed carry out approximate calculation, complete the update processing of check-node；
Step 3: the channel massage of the belief messages and reception transmitted to variable node by check-node, calculates variable section
Point arrives the newer belief messages of check-node, completes the update processing of variable node；
Step 4: the channel massage of the belief messages and reception transmitted to variable node by check-node, calculates variable section
The belief messages of point；
Step 5: the belief messages to variable node carry out hard decision, a sequence isometric with code length is obtained, if described
Sequence meets the parity check equation of LDPC code or reaches maximum iteration, then exports sequence after LDPC code decoding, no
Then return to step two, until obtained sequence meets the parity check equation of LDPC code or reaches maximum iteration.
2. according to the method described in claim 1, it is characterized in that, all variable nodes of initialization are transmitted to check-node
Belief messages, specifically include：
Pass through posterior probability PiIt calculates, the confidence that all variable nodes are transmitted to check-node is initialized and stored with its result
Spend message Lvc, to some information bit i, posterior probabilityWherein, Pr { zi=b | yi},b∈{0,1}
Y is exported for given channeliUnder the conditions of, code word bits z after LDPC codingsiPosterior probability equal to b, when the mean value of channel is 0, side
Difference is σ2Awgn channel, then
3. according to the method described in claim 2, it is characterized in that, the confidence transmitted by variable node to check-node
It spends and passes through the revised minimum value of modifying factor in message, and the sub-minimum obtained by the revised estimation of modifying factor,
Approximate evaluation is carried out to the belief messages that check-node to variable node transmits, completes the update processing of check-node, specifically
Including：
To each check-node c according to formula：
Transmit message RcvTo change
Node v is measured, wherein：Normalize modifying factor
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvFor the soft letter received
Breath, σ2For the variance of interchannel noise.
4. according to the method described in claim 2, it is characterized in that, the confidence transmitted by variable node to check-node
It spends and passes through the revised minimum value of modifying factor in message, and the sub-minimum obtained by the revised estimation of modifying factor,
Approximate evaluation is carried out to the belief messages that check-node to variable node transmits, completes the update processing of check-node, specifically
Including：
To each check-node c according to formula：
Transmit message RcvTo change
Node v is measured, wherein：
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvFor the soft letter received
Breath, σ2For the variance of interchannel noise.
5. according to the method described in claim 2, it is characterized in that, the confidence transmitted by variable node to check-node
It spends and passes through the revised minimum value of modifying factor in message, and the sub-minimum obtained by the revised estimation of modifying factor,
Approximate evaluation is carried out to the belief messages that check-node to variable node transmits, completes the update processing of check-node, specifically
Including：
To each check-node c according to formula
Calculate each verification
Node c passes to the belief messages of variable node v, wherein：
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvFor the soft letter received
Breath, σ2For the variance of interchannel noise.
6. according to the method described in any one of claim 3-5, which is characterized in that passed to variable node by check-node
The channel massage of the belief messages and reception passed calculates the belief messages of variable node, specifically includes：
According toCalculate the belief messages of variable node.
7. according to the method described in claim 6, it is characterized in that, carry out hard decision to the belief messages of variable node, obtain
The sequence isometric with code length to one, if the sequence meets the parity check equation of LDPC code or reaches greatest iteration time
Number then exports sequence after LDPC code decoding, otherwise return to step two, until obtained sequence meets the even-odd check of LDPC code
Equation or until reaching maximum iteration, specifically includes：
Hard decision is carried out to the belief messages of variable node, obtains a sequence isometric with code length, if the sequence is full
The parity check equation of sufficient LDPC code reaches maximum iteration, then sequence after LDPC code decoding is exported, otherwise to becoming
After the minimum value for the belief messages that amount node is transmitted to check-node re-starts amendment, and variable node is calculated again
Belief messages, a sequence isometric with code length is obtained after hard decision, until obtained sequence meets the odd even of LDPC code
Check equations or until reaching maximum iteration.
8. the method according to the description of claim 7 is characterized in that the belief messages to variable node carry out hard decision, obtain
The sequence isometric with code length to one, if the sequence meets the parity check equation of LDPC code or reaches greatest iteration time
Number then exports sequence after LDPC code decoding, the minimum value of the belief messages otherwise transmitted to variable node to check-node
After re-starting amendment, and the belief messages of variable node are calculated again, obtained after hard decision one it is isometric with code length
Sequence, it is specific to wrap until obtained sequence meets the parity check equation of LDPC code or reaches maximum iteration
It includes：
If Lv＜ 0, the then code word bits for decoding outOtherwiseIfOr reach iteration time
L=Itermax is counted, then iteration stopping, otherwise iterations l=l+1, the confidence level transmitted to variable node to check-node disappear
After the minimum value of breath re-starts amendment, and the belief messages of variable node are calculated again, one is obtained after hard decision
The isometric sequence with code length is until obtained sequence meets the parity check equation of LDPC code or reaches maximum iteration
Only.
9. a kind of multiple-factor based on parameter Estimation corrects LDPC code code translator, which is characterized in that including：
Initialization unit initializes the confidence that all variable nodes are transmitted to check-node for the channel information by receiving
Spend message；
Processing unit is updated, modifying factor amendment is passed through in the belief messages for being transmitted to check-node by variable node
Minimum value afterwards, and the sub-minimum of belief messages is obtained by the revised estimation of modifying factor, to check-node to change
It measures the belief messages that node transmits and carries out approximate evaluation estimation, complete the update processing of check-node；Pass through check-node pair
The channel massage of belief messages and reception that variable node transmits calculates variable node and disappears to the newer confidence level of check-node
Breath completes the update processing of variable node；Pass through the channel of belief messages and reception that check-node transmits variable node
Message calculates the belief messages of variable node；
Decision process unit carries out hard decision to the belief messages of variable node, obtains a sequence isometric with code length, such as
Sequence described in fruit meets the parity check equation of LDPC code or reaches maximum iteration, then exports LDPC code decoding postorder
Row, otherwise iterations add one, and restart to calculate, until obtained sequence meet LDPC code parity check equation or
Until reaching maximum iteration.
10. device according to claim 9, which is characterized in that
The decision process unit is additionally operable to, to each check-node c according to formula：
Transmit message RcvTo change
Node v is measured, wherein：Normalize modifying factor
Alternatively,
To each check-node c according to formula：
Transmit message RcvTo change
Node v is measured, wherein：
Alternatively,
To each check-node c according to formula
Wherein：
According toEach variable node v transmits LvcCheck-node c is given, wherein：rvFor the soft letter received
Breath, σ2For the variance of interchannel noise；According toThe belief messages for calculating variable node, if Lv＜ 0,
The code word bits for then decoding outOtherwiseIfOr reach iterations l=Itermax,
Then iteration stopping, otherwise l=l+1 are then back to the first step, and by revised code word bitsOutput.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810342373.5A CN108494412A (en) | 2018-04-17 | 2018-04-17 | A kind of multiple-factor amendment LDPC code interpretation method and device based on parameter Estimation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810342373.5A CN108494412A (en) | 2018-04-17 | 2018-04-17 | A kind of multiple-factor amendment LDPC code interpretation method and device based on parameter Estimation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108494412A true CN108494412A (en) | 2018-09-04 |
Family
ID=63316221
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810342373.5A Pending CN108494412A (en) | 2018-04-17 | 2018-04-17 | A kind of multiple-factor amendment LDPC code interpretation method and device based on parameter Estimation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108494412A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109361491A (en) * | 2018-09-26 | 2019-02-19 | 西安科技大学 | Interpretation method, equipment and the storage medium of multi-input multi-output system based on LDPC |
| CN109586844A (en) * | 2018-10-30 | 2019-04-05 | 北京航空航天大学 | A kind of nonuniform protection interpretation method and system based on set |
| CN110830050A (en) * | 2019-11-27 | 2020-02-21 | 武汉虹信通信技术有限责任公司 | LDPC decoding method, system, electronic device and storage medium |
| CN112003626A (en) * | 2020-08-31 | 2020-11-27 | 武汉梦芯科技有限公司 | LDPC decoding method, system and medium based on known bits of navigation message |
| CN112039533A (en) * | 2020-08-05 | 2020-12-04 | Oppo广东移动通信有限公司 | Data processing method, device, decoder, equipment and computer storage medium |
| CN112260698A (en) * | 2019-07-22 | 2021-01-22 | 上海高清数字科技产业有限公司 | Dynamic correction factor configuration method in LDPC decoder |
| CN112350736A (en) * | 2019-07-22 | 2021-02-09 | 上海高清数字科技产业有限公司 | Dynamic correction factor configuration method in LDPC decoder |
| CN113014267A (en) * | 2019-12-18 | 2021-06-22 | 华为技术有限公司 | Decoding method, apparatus, readable storage medium, chip and computer program product |
| CN113055028A (en) * | 2021-03-18 | 2021-06-29 | 北京得瑞领新科技有限公司 | LDPC decoding method, decoder, decoding device and storage medium |
| CN113271111A (en) * | 2021-06-03 | 2021-08-17 | 北京邮电大学 | Decoding method and system based on improved minimum sum algorithm |
| CN113300721A (en) * | 2021-05-27 | 2021-08-24 | 北京航空航天大学合肥创新研究院(北京航空航天大学合肥研究生院) | LDPC code decoding method and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101345532A (en) * | 2008-08-15 | 2009-01-14 | 苏州大学 | Decoding method for LDPC channel code |
| CN103199874A (en) * | 2012-01-05 | 2013-07-10 | 国民技术股份有限公司 | Low density parity check code decoding method |
| CN105991141A (en) * | 2015-02-15 | 2016-10-05 | 上海高清数字科技产业有限公司 | LDPC decoding method |
-
2018
- 2018-04-17 CN CN201810342373.5A patent/CN108494412A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101345532A (en) * | 2008-08-15 | 2009-01-14 | 苏州大学 | Decoding method for LDPC channel code |
| CN103199874A (en) * | 2012-01-05 | 2013-07-10 | 国民技术股份有限公司 | Low density parity check code decoding method |
| CN105991141A (en) * | 2015-02-15 | 2016-10-05 | 上海高清数字科技产业有限公司 | LDPC decoding method |
Non-Patent Citations (1)
| Title |
|---|
| 肖婧婷: "《低复杂度LDPC码编译码算法研究与实现》", 《中国学术期刊（光盘版）电子杂志社》 * |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109361491A (en) * | 2018-09-26 | 2019-02-19 | 西安科技大学 | Interpretation method, equipment and the storage medium of multi-input multi-output system based on LDPC |
| CN109586844A (en) * | 2018-10-30 | 2019-04-05 | 北京航空航天大学 | A kind of nonuniform protection interpretation method and system based on set |
| CN109586844B (en) * | 2018-10-30 | 2020-08-04 | 北京航空航天大学 | Set-based unequal protection decoding method and system |
| CN112260698A (en) * | 2019-07-22 | 2021-01-22 | 上海高清数字科技产业有限公司 | Dynamic correction factor configuration method in LDPC decoder |
| CN112350736A (en) * | 2019-07-22 | 2021-02-09 | 上海高清数字科技产业有限公司 | Dynamic correction factor configuration method in LDPC decoder |
| CN110830050B (en) * | 2019-11-27 | 2023-09-29 | 武汉虹旭信息技术有限责任公司 | LDPC decoding method, system, electronic equipment and storage medium |
| CN110830050A (en) * | 2019-11-27 | 2020-02-21 | 武汉虹信通信技术有限责任公司 | LDPC decoding method, system, electronic device and storage medium |
| CN113014267A (en) * | 2019-12-18 | 2021-06-22 | 华为技术有限公司 | Decoding method, apparatus, readable storage medium, chip and computer program product |
| CN113014267B (en) * | 2019-12-18 | 2024-04-23 | 华为技术有限公司 | Decoding method, device, readable storage medium, chip and computer program product |
| CN112039533A (en) * | 2020-08-05 | 2020-12-04 | Oppo广东移动通信有限公司 | Data processing method, device, decoder, equipment and computer storage medium |
| CN112039533B (en) * | 2020-08-05 | 2023-10-10 | Oppo广东移动通信有限公司 | Data processing method, device, decoder, equipment and computer storage medium |
| CN112003626A (en) * | 2020-08-31 | 2020-11-27 | 武汉梦芯科技有限公司 | LDPC decoding method, system and medium based on known bits of navigation message |
| CN112003626B (en) * | 2020-08-31 | 2023-11-10 | 武汉梦芯科技有限公司 | LDPC decoding method, system and medium based on navigation message known bits |
| CN113055028A (en) * | 2021-03-18 | 2021-06-29 | 北京得瑞领新科技有限公司 | LDPC decoding method, decoder, decoding device and storage medium |
| CN113300721B (en) * | 2021-05-27 | 2022-11-22 | 北京航空航天大学合肥创新研究院(北京航空航天大学合肥研究生院) | LDPC code decoding method and storage medium |
| CN113300721A (en) * | 2021-05-27 | 2021-08-24 | 北京航空航天大学合肥创新研究院(北京航空航天大学合肥研究生院) | LDPC code decoding method and storage medium |
| CN113271111A (en) * | 2021-06-03 | 2021-08-17 | 北京邮电大学 | Decoding method and system based on improved minimum sum algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108494412A (en) | A kind of multiple-factor amendment LDPC code interpretation method and device based on parameter Estimation | |
| CN108282264B (en) | Polar code decoding method based on bit flipping serial elimination list algorithm | |
| CN105720992B (en) | A kind of simplified decoding method of polarization code | |
| JP2533219B2 (en) | Viterbi decoding method and apparatus | |
| CN103259545B (en) | Quasi-cyclic low density odd-even check code belief propagation decoding method based on oscillation | |
| US20090222711A1 (en) | Generalized Multi-Threshold Decoder for Low-Density Parity Check Codes | |
| CN101707485A (en) | LDPC decoding method combining bit flipping (BF) and majority logic (MLG) | |
| CN110100403A (en) | The error detection in communication system transmitted using Polarization Coding data | |
| CN111342935B (en) | High-performance soft-decision decoding method based on QC-LDPC code | |
| CN101345601A (en) | Interpretation method and decoder | |
| CN107453807A (en) | A kind of polarization method, device and the electronic equipment of atmospheric optical communication channel model | |
| CN103973314A (en) | Signal coding and decoding method based on LDPC, receiving end and sending end | |
| CN101273531B (en) | Improved TURBO-decoding information transmission method, apparatus and system of low density parity check code | |
| CN105812000B (en) | A kind of improved BCH soft-decision decoding method | |
| CN110995279B (en) | Polarization code combined SCF spherical list overturning decoding method | |
| CN114070331B (en) | Self-adaptive serial offset list flip decoding method and system | |
| CN106209312A (en) | A kind of cyclic code parameter blind recognition algorithm utilizing soft-decision | |
| CN110430153A (en) | The frequency deviation modification method and device of satellite communication | |
| CN106656209A (en) | Cascaded code method adopting iterative decoding for correcting synchronization errors | |
| CN106027437A (en) | Carrier synchronization method and system in helicopter satellite communication | |
| CN104184480B (en) | A kind of LDPC for reducing complexity improves interpretation method | |
| CN106301389B (en) | Interpretation method and equipment | |
| CN112491422A (en) | Gauss optimization-based bit flipping serial elimination list decoding method and system | |
| CN109995383A (en) | Interpretation method, device, storage medium and computer program product | |
| CN108667593B (en) | FPGA-based time diversity parallel synchronization method for resisting helicopter rotor wing shielding |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180904 |