CN104779962B - The minimum segment vectors of Max Log MAP decoding algorithm complexities efficiently produce method - Google Patents
The minimum segment vectors of Max Log MAP decoding algorithm complexities efficiently produce method Download PDFInfo
- Publication number
- CN104779962B CN104779962B CN201510154848.4A CN201510154848A CN104779962B CN 104779962 B CN104779962 B CN 104779962B CN 201510154848 A CN201510154848 A CN 201510154848A CN 104779962 B CN104779962 B CN 104779962B
- Authority
- CN
- China
- Prior art keywords
- minimal
- trellises
- segment vectors
- depth
- vetsec
- 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.)
- Expired - Fee Related
Links
Landscapes
- Error Detection And Correction (AREA)
Abstract
Method is efficiently produced the invention discloses a kind of minimum segment vectors of Max Log MAP decoding algorithm complexities, is comprised the following steps：1) according to the minimal trellises M of Max Log MAP decoding algorithmsmin, obtain the minimal trellises MminState vector corresponding to upper each depth and information bit vector；2) according to the minimal trellises M of step 1)minState vector and information bit vector obtain minimal trellises M corresponding to upper each depthminIn each section of segment vectors；3) combination step 2) obtained minimal trellises MminIn each section of segment vectors, obtain for the minimum segment vectors Vetsec of Max Log MAP decoding algorithm computation complexities.The present invention can efficiently generate the minimum segment vectors Vetsec of Max Log MAP decoding algorithm computation complexities.
Description
Technical field
The invention belongs to communication technical field, is related to a kind of segment vectors Vetsec generation method, and in particular to a kind of
The minimum segment vectors of Max-Log-MAP decoding algorithm complexities efficiently produce method.
Background technology
In base band transmission, channel coding (error correcting code) often consumes a big chunk energy of communication equipment, such as exists
B.Bougard et al.,“Energy-scalability enhancement of wireless local area
network transceivers,”in Proc.2004IEEE Workshop on Signal Processing Advances
In in Wireless Communications, pp.449-453, Viterbi decoding in typical 802.11 receiver,
Occupy the 35% of its total power consumption.The problem of energy expenditure, governs communication equipment miniaturization and mobility, how to improve
The code check of coding so that the generation code word of unit length can transmit more information bits, have very big application value.For
For Turbo code decoding, total bitrate is improved, typically takes punctured mode, such as M.A.Kousa and A.H.Mugaibel,
“Puncturing effects on turbo codes,”IEEE Proc.Commun.,vol.149,no.3,pp.132-
Method described in 138, June 2002.Another method for improving total bitrate is member's code using high code check, for
For convolutional code C (n, k, v), this method has the percent of pass improved in the unit interval, reduces the effect such as time delay, but with k
Increase, the complexity of decoding can exponentially increase again with k increase.
Convolutional code C (n, k, v) can represent that this grid chart is the cycle with the grid chart of a semo-infinite, it
The most short cycle is referred to as mesh module.For the mesh module that a code check is k/n, it containsIndividual segmentation, have in depth iIndividual state, each state are launchedBar side, ai(0≤i≤n-1) represents connection depth i a certain state and depth i+1
A certain state side representated by output codons bit number, total constraint length is v, viAnd biIt is known respectively as the grid
State complexity and branching complexity of the module on depth i, for Viterbi algorithm, the complexity of grid module
It is defined as：
Conventional mesh figure (Conventional trellis) is a kind of performance shape of the most frequently used convolutional code mesh module
Formula, it containsIndividual segmentation, there is 2vIndividual original state and 2vIndividual final state, each original state exhale 2kBar side with
Final state is connected, and each edge represents n-bit (see Fig. 1).
Another important convolutional code mesh module takes the form of minimal trellises (Minimal trellis), and it contains
HaveIndividual segmentation, it is information (b for k therein segmentationi=1), other n-k segmentation is asemantic
(bi=0), minimal trellises contain 2vIndividual original state and 2vIndividual final state, each edge represent 1 bit generation code word (see
Fig. 2).Minimal trellises can enter row decoding in Viterbi decoding in the minimum method of complexity, therefore apply minimal trellises,
Corresponding energy expenditure can be reduced and improve the utilization rate of hardware.But translated for other decoded modes, such as Max-Log-MAP
Code, minimal trellises do not ensure that its complexity is necessarily minimum.
State on a certain depth i of minimal trellises is all removed, and by the state on depth i-1 and depth i+1
State be joined directly together, thus obtained a new segmentation, it is closed by two on original minimal trellises segmentations
And form, and the grid chart after merging only is leftIndividual segmentation.Popularization is come, this that multiple segmentations are merged into one
The operation of segmentation is referred to as staged operation, and the grid chart obtained after segmentation is referred to as being segmented grid chart (Sectionalized
trellis).Staged operation is a kind of applied to, to obtain the operation of new grid graph topological structure, it is logical on minimal trellises
The state deleted on a certain depth i of minimal trellises is crossed, and the state on depth i-1 and the state on depth i+1 are entered again
Row connection, obtains an overall compact structure.More and more with the number of segmentation, grid graph structure will be further compact,
Finally, after intermediateness is all deleted, conventional mesh figure is obtained.For convolutional code C (n, k, v), its possible point
Segmented mode shares 2n-1Kind, in order to express a variety of segmented modes for minimal trellises, we define a binary system
Segment vectors vetsec={ vetseci, i=1 ..., n-1, when minimal trellises depth i state be deleted and incite somebody to action
After state on depth i-1 is joined directly together with the state on depth i+1, then vetseci=1, otherwise, vetseci=0 (see Fig. 3).
We use respectivelyTo represent to be segmented the ratio of the generation code word representated by side of the grid chart on depth i to depth i+1
Special number and state and the complexity of branch on depth i,For example, Fig. 3 is the C (5,3,5) shown in Fig. 2
Segmentation grid chart of the minimal trellises under conditions of segment vectors vetsec=(0010), in the minimal trellises shown in Fig. 2
Staged operation is carried out in depth 3, i.e., is deleted the state in depth 3, and depth 2 and the state in depth 4 are joined directly together.Carry out
Grid chart after staged operation sharesIndividual segmentation, its side complexity vectorAnd
The trellis complexity defined by (1) formula, it can be deduced that the complexity TC (M of conventional mesh figureconv)=427.5, minimum grid
Complexity TC (the M of figuremin)=74.7, and under conditions of segment vectors vetsec=(0010), it is segmented the complexity of grid chart
TC(Msec)=96.
In R.J.McEliece and W.Lin, " The trellis complexity of convolutional
Codes, " grid chart defined in IEEE Trans.Inf.Theory, vol.42, no.6, pp.1855-1864, Nov.1996
Theoretical complexity in, for Viterbi decoding for, the complexity of minimal trellises is less than the complexity of conventional mesh figure
Degree, and can't bring any performance loss with minimal trellises.But for Max-Log-MAP decoding algorithms, fortune
Minimum decoding complexity will can not necessarily be obtained by entering row decoding with minimal trellises.
Each section of addition, multiplication and ratio can be used by being segmented the computation complexity of the Max-Log-MAP decoding algorithms of grid chart
Compared with the cumulative of number represent that here we represent addition, multiplication and comparison operation with M, S, C respectively.In Moritz G,
Souza R,Pimentel C,et al.“Turbo Decoding Using the Sectionalized Minimal
Trellis of the Constituent Code:Provided in Performance-Complexity Trade-Off " .2013.
Max-Log-MAP decoding algorithms are for γ in i-th of the segmentation of segmentation grid charti,αi,βi,ΛiComputation complexity and total
The calculation formula of computation complexity.By taking C (5,3,5) as an example, suitable packet mode is selected in Max-Log-MAP decodings,
Under conditions of losing certain performance, the addition of the minimum segmented mode of its computation complexity and the number of comparison operation and traditional net
Compared to 63.8% and 59.4% is reduced respectively, complexity greatly reduces trrellis diagram.How existing technical scheme is for find C
The segment vectors that the Max-Log-MAP decoding algorithm complexities of (n, k, v) code are minimum do not have special method, can only
To its 2n-1Kind of segmented mode compare after traversal computing drawing.
The content of the invention
The effect of the present invention is to overcome the shortcomings that above-mentioned existing method need to carry out traversal computing, there is provided a kind of Max-
The minimum segment vectors of Log-MAP decoding algorithm complexities efficiently produce method, and this method can be generated efficiently
The minimum segment vectors Vetsec of Max-Log-MAP decoding algorithm computation complexities.
To reach above-mentioned purpose, the minimum segmentation of Max-Log-MAP decoding algorithms computation complexity of the present invention to
Amount Vetsec generation method comprises the following steps：
1) according to the minimal trellises M of Max-Log-MAP decoding algorithmsmin, obtain the minimal trellises MminUpper each depth
State vector corresponding to degree and information bit vector；
2) according to the minimal trellises M of step 1)minState vector and information bit vector obtain most corresponding to upper each depth
Small grid figure MminIn each section of segment vectors；
3) combination step 2) obtained minimal trellises MminIn each section of segment vectors, obtain translating for Max-Log-MAP
The minimum segment vectors Vetsec of code algorithm computation complexity.
The detailed process of step 2) is：
Judge minimal trellises MminState vector v corresponding on upper depth iiAnd information bit vector biWith on depth i+1
Corresponding state vector vi+1Size, wherein, i=0,1 ..., n-2, n be minimal trellises MminThe sum of upper depth；
Work as vi=vi+1When -1, then minimal trellises MminThe segment vectors vetsec of middle i+1 sectioni+1=1；
Work as vi=vi+1When+1, then minimal trellises MminThe segment vectors vetsec of middle i+1 sectioni+1=0；
Work as vi=vi+1, and minimal trellises MminInformation bit vector b corresponding on upper depth iiWhen=1, then minimum grid
Scheme MminThe segment vectors vetsec of middle i+1 sectioni+1=0；
Work as vi=vi+1, and minimal trellises MminInformation bit vector b corresponding on upper depth iiWhen=0, then minimum grid
Scheme MminThe segment vectors vetsec of middle i+1 sectioni+1=1.
The invention has the advantages that：
The side of efficiently producing of the minimum segment vectors of Max-Log-MAP decoding algorithms complexity of the present invention
Method, only need to be according to its minimal trellises M in operationminState vector v corresponding to upper each depthiAnd information bit vector biRatio
Minimal trellises M is obtained compared with judgingminIn each section of segment vectors vetseci, then each section of segment vectors are combined i.e.
Can, compared with the minimum segment vectors Vetsec of tradition generation complexity needs to carry out traveling through computing, the present invention improves searching
The efficiency of Max-Log-MAP decoding algorithm computation complexity lowest segment modes.In addition, in theory, for there is time-varying
The application scenario of generator matrix, under conditions of its minimal trellises is known, this method can quick-pick be answered in real time online
The minimum component vector paragraph of miscellaneous degree.
Brief description of the drawings
Fig. 1 is the conventional mesh figure of C (5,3,5) code；
Fig. 2 is the minimal trellises of C (5,3,5) code；
Fig. 3 is segmentation grid chart of C (5,3,5) codes under conditions of segment vectors vetsec=(0010)；
Fig. 4 is C (7,4,4) minimal trellises；
Fig. 5 is the complicated dynamic behaviour result figure under C (7,4,4) the various segmented modes being calculated with traversal.
Embodiment
The present invention is described in further detail below in conjunction with the accompanying drawings：
The side of efficiently producing of the minimum segment vectors of Max-Log-MAP decoding algorithms complexity of the present invention
Method comprises the following steps：
1) according to the minimal trellises M of Max-Log-MAP decoding algorithmsmin, obtain the minimal trellises MminUpper each depth
State vector corresponding to degree and information bit vector；
2) according to the minimal trellises M of step 1)minState vector and information bit vector obtain most corresponding to upper each depth
Small grid figure MminIn each section of segment vectors；
3) combination step 2) obtained minimal trellises MminIn each section of segment vectors, obtain translating for Max-Log-MAP
The minimum segment vectors Vetsec of code algorithm computation complexity.
The detailed process of step 2) is：
Judge minimal trellises MminState vector v corresponding on upper depth iiAnd information bit vector biWith on depth i+1
Corresponding state vector vi+1Size, wherein i=0,1 ..., n-2, n be minimal trellises MminThe sum of upper depth；
Work as vi=vi+1When -1, then minimal trellises MminThe segment vectors vetsec of middle i+1 sectioni+1=1；
Work as vi=vi+1When+1, then minimal trellises MminThe segment vectors vetsec of middle i+1 sectioni+1=0；
Work as vi=vi+1, and minimal trellises MminInformation bit vector b corresponding to upper depth iiWhen=1, then minimal trellises
MminThe segment vectors vetsec of middle i+1 sectioni+1=0；
Work as vi=vi+1, and minimal trellises MminInformation bit vector b corresponding to upper depth iiWhen=0, then minimal trellises
MminThe segment vectors vetsec of middle i+1 sectioni+1=1.
For C (n, k, v), segmented decodings are carried out to it with Max-Log-MAP algorithms, if necessary to choose complexity most
Low segmented mode, it is necessary to its 2n-1Kind segmented mode is traveled through, adding required for calculating respectively under every kind of segmented mode
Method, multiplication and the number compared.And the method used in the present invention, it is only necessary to knowing the state vector V=of its minimal trellises
{viAnd information bit vector B={ biIn the case of, it is possible to draw minimum one of its complexity by simply adjudicating computing
Kind of segmented mode, it is secondary no more than 2* (n-1) to adjudicate the total degree of computing, and uses traditional traversal computing to needSecondary computing, wherein WithIt is illustrated respectively in
The γ, α, β for the i+1 section being segmented under jth kind segmented mode in grid chart, Λ computation complexity and total segments, this is right
It is significant in finding out the amount of calculation that the minimum Max-Log-MAP decoding algorithms of computation complexity are saved rapidly, and with n, k
Increase, the advantage of this method can be further obvious.In addition, for there is the application scenario of time-varying generator matrix, knowing it most
Under conditions of small grid figure, this method can the minimum component vector paragraph of quick-pick complexity in real time online, greatly
Improve the speed for finding Max-Log-MAP decoding algorithm computation complexity lowest segment modes.This computation complexity is minimum
The determination of segmented mode is applied to the application scenario sensitive to computational complexity, is reducing the energy consumption of communication equipment, is improving computing
It is considerable in terms of efficiency, by taking C (7,4,4) code shown in Fig. 4 as an example, with above-mentioned generation method：
v0=v1- 1, so minimal trellises MminIn the 1st section of segment vectors vetsec1=1；
v1=v2- 1, so minimal trellises MminIn the 2nd section of segment vectors vetsec2=1；
v2=v3+ 1, so minimal trellises MminIn the 3rd section of segment vectors vetsec3=0；
v3=v4, and minimal trellises MminInformation bit vector b corresponding to upper depth 33When=1, so minimal trellises
MminIn the 4th section of segment vectors vetsec4=0；
v4=v5+ 1, so minimal trellises MminIn the 5th section of segment vectors vetsec5=0.
v5=v6- 1, so minimal trellises MminIn the 6th section of segment vectors vetsec6=1；
The minimum segmented mode vetsec=[110001] of the computation complexity that is obtained according to above-mentioned generation method, with Fig. 5
Shown by traveling through the minimum segmented mode vetsec=[110001] of computation complexity that computing obtains is the same, and it is transported
Calculating number saving rate isAbout 99.99%.
In Fig. 5, S, M, C represent addition, multiplication and comparison operation number respectively, and Tr, Ta, Tb, Tv represent corresponding respectively
Segmented mode under calculate α, β, γ, the operand needed for Λ in Max-Log-MAP decoding algorithms.
Note：Fig. 1 to Fig. 3 selects from Moritz G L, Demo Souza R, Pimentel C, et al.Turbo
decoding using the sectionalized minimal trellis of the constituent code:
performance-complexity trade-off[J].Communications,IEEE Transactions on,2013,
61(9):3600-3610. Fig. 4 selects from Benchimol I, Pimentel C, Demo Souza R.Sectionalization
of the minimal trellis module for convolutional codes[C]//Telecommunications
and Signal Processing(TSP),201235th International Conference on.IEEE,2012:
227-232。
Claims (1)
1. a kind of minimum segment vectors of Max-Log-MAP decoding algorithms complexity efficiently produce method, its feature
It is, comprises the following steps：
1) according to the minimal trellises M of Max-Log-MAP decoding algorithmsmin, obtain the minimal trellises MminUpper each depth pair
State vector and the information bit vector answered；
2) according to the minimal trellises M of step 1)minState vector and information bit vector obtain minimum net corresponding to upper each depth
Trrellis diagram MminIn each section of segment vectors；
3) combination step 2) obtained minimal trellises MminIn each section of segment vectors, obtain decoding for Max-Log-MAP and calculate
The minimum segment vectors Vetsec of method computation complexity；
The detailed process of step 2) is：
Judge minimal trellises MminState vector v corresponding on upper depth iiAnd information bit vector biIt is corresponding with depth i+1
State vector vi+1Size, wherein, i=0,1 ..., n-2, n be minimal trellises MminThe sum of upper depth；
Work as vi=vi+1When -1, then minimal trellises MminThe segment vectors vetsec of middle i+1 sectioni+1=1；
Work as vi=vi+1When+1, then minimal trellises MminThe segment vectors vetsec of middle i+1 sectioni+1=0；
Work as vi=vi+1, and minimal trellises MminInformation bit vector b corresponding on upper depth iiWhen=1, then minimal trellises
MminThe segment vectors vetsec of middle i+1 sectioni+1=0；
Work as vi=vi+1, and minimal trellises MminInformation bit vector b corresponding on upper depth iiWhen=0, then minimal trellises
MminThe segment vectors vetsec of middle i+1 sectioni+1=1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510154848.4A CN104779962B (en) | 2015-04-02 | 2015-04-02 | The minimum segment vectors of Max Log MAP decoding algorithm complexities efficiently produce method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510154848.4A CN104779962B (en) | 2015-04-02 | 2015-04-02 | The minimum segment vectors of Max Log MAP decoding algorithm complexities efficiently produce method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104779962A CN104779962A (en) | 2015-07-15 |
| CN104779962B true CN104779962B (en) | 2018-03-02 |
Family
ID=53621237
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510154848.4A Expired - Fee Related CN104779962B (en) | 2015-04-02 | 2015-04-02 | The minimum segment vectors of Max Log MAP decoding algorithm complexities efficiently produce method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104779962B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6973615B1 (en) * | 2000-12-15 | 2005-12-06 | Conexant Systems, Inc. | System of and method for decoding trellis codes |
| CN101286817B (en) * | 2008-04-03 | 2011-07-20 | 浙江大学 | General decoding method for conventional binary and double-binary Turbo code |
| CN102611464A (en) * | 2012-03-30 | 2012-07-25 | 电子科技大学 | Turbo decoder based on external information parallel update |
| US8358713B2 (en) * | 2007-09-10 | 2013-01-22 | Sarath Babu Govindarajulu | High throughput and low latency map decoder |
| CN103200142A (en) * | 2013-03-22 | 2013-07-10 | 西安电子科技大学 | Two-state simplified method of non-recursive shaped offset quadrature phase shift keying (SOQPSK)-TG signal |
-
2015
- 2015-04-02 CN CN201510154848.4A patent/CN104779962B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6973615B1 (en) * | 2000-12-15 | 2005-12-06 | Conexant Systems, Inc. | System of and method for decoding trellis codes |
| US8358713B2 (en) * | 2007-09-10 | 2013-01-22 | Sarath Babu Govindarajulu | High throughput and low latency map decoder |
| CN101286817B (en) * | 2008-04-03 | 2011-07-20 | 浙江大学 | General decoding method for conventional binary and double-binary Turbo code |
| CN102611464A (en) * | 2012-03-30 | 2012-07-25 | 电子科技大学 | Turbo decoder based on external information parallel update |
| CN103200142A (en) * | 2013-03-22 | 2013-07-10 | 西安电子科技大学 | Two-state simplified method of non-recursive shaped offset quadrature phase shift keying (SOQPSK)-TG signal |
Non-Patent Citations (3)
| Title |
|---|
| urbo Decoding Using the Sectionalized Minimal Trellis of the Constituent Code: Performance-Complexity Trade-Off;Guilherme Luiz Moritz，et al;《IEEE Transactions on Communications》;20130930;第61卷(第9期);第3600-3609页 * |
| 低复杂度Log-MAP译码算法的研究;毕岗 等;《计算机工程与应用》;20111231;第47卷(第10期);第89-91,97页 * |
| 高性能的Max-Log-MAP线性分段算法研究;毕岗 等;《电路与系统学报》;20121231;第17卷(第6期);第27-30页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104779962A (en) | 2015-07-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101388674B (en) | Decoding method, decoder and Turbo code decoder | |
| CN104079382B (en) | A kind of polarization code decoder based on probability calculation and polarization code coding method | |
| CN108847850A (en) | A kind of segmentation polarization code coding/decoding method based on CRC-SSCL | |
| CN105515590B (en) | A kind of effective low complex degree serially offsets list polarization code coding method | |
| CN105634507A (en) | Assembly-line architecture of polarization code belief propagation decoder | |
| CN108039891A (en) | A kind of polarization code BP interpretation methods and device based on multistage more new technological process | |
| CN107231158A (en) | A kind of polarization code iterative receiver, system and polarization code iterative decoding method | |
| CN110138390A (en) | A kind of polarization code SSCL algorithm decoder based on deep learning | |
| CN112332864A (en) | Polar code decoding method and system for self-adaptive ordered mobile pruning list | |
| CN102904667A (en) | Method for decoding tail biting convolution codes of PBCH (physical broadcast channel) decoding in LTE (long term evolution) | |
| CN101262232A (en) | A decoding algorithm for overlapping coding multiplexing | |
| CN103124181A (en) | Turbo code decoding iteration cease method based on cosine similarity | |
| CN106656216A (en) | Modified soft-input soft-output decoding method for Turbo product codes | |
| CN103986557A (en) | LTE Turbo code parallel block decoding method with low path delay | |
| CN102611464B (en) | Turbo decoder based on external information parallel update | |
| CN102832954B (en) | Turbo code iterative decoding stopping method based on soft information average minimum value | |
| CN108462561A (en) | The channel decoding method and device gone here and there and combined in ultrahigh speed communication system | |
| CN109639290B (en) | Semi-random grouping superposition coding and decoding method | |
| CN104779962B (en) | The minimum segment vectors of Max Log MAP decoding algorithm complexities efficiently produce method | |
| CN102377438B (en) | Channel decoding method and tail biting convolutional decoder | |
| Chen et al. | Design of a low power viterbi decoder for wireless communication applications | |
| CN103701475B (en) | Decoding method for Turbo codes with word length of eight bits in mobile communication system | |
| CN105610550B (en) | A kind of Viterbi interpretation method for power line carrier communication | |
| CN103944673A (en) | Low-delay progressive decoding method suitable for rateless codes | |
| Qian et al. | A depth-first ML decoding algorithm for tail-biting trellises |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180302 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |