WO2015170949A1 - Generalised decoder employing belief propagation using codes other than ldpc codes - Google Patents
Generalised decoder employing belief propagation using codes other than ldpc codes Download PDFInfo
- Publication number
- WO2015170949A1 WO2015170949A1 PCT/MA2014/000026 MA2014000026W WO2015170949A1 WO 2015170949 A1 WO2015170949 A1 WO 2015170949A1 MA 2014000026 W MA2014000026 W MA 2014000026W WO 2015170949 A1 WO2015170949 A1 WO 2015170949A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- variables
- level
- test values
- codes
- interleaved
- Prior art date
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
- H03M13/1191—Codes on graphs other than LDPC codes
-
- 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
- H03M13/29—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 combining two or more codes or code structures, e.g. product codes, generalised product codes, concatenated codes, inner and outer codes
- H03M13/2906—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 combining two or more codes or code structures, e.g. product codes, generalised product codes, concatenated codes, inner and outer codes using block codes
Definitions
- the present invention relates to a decoder for error correction code that extends the decoders based on "BeliefPropagation" or propagation of confidence. This decoder minimizes the limitations of the latter by no longer being limited to hollow matrices for the coding of information.
- Decoder based on extended confidence propagation, can improve the decoding performance even for small or medium size hollow coding matrices and even for non-hollow matrix. It allows easier hardware implementation than conventional decoders such as LDPCs.
- Digital communications are based on a three-part transmission model. The first part is the transmitter, the second is the transmission channel, and the third part is the receiver, see Figure 01.
- the transmission channel is always subject to disturbances or noises. These disturbances induce errors in the information to be transmitted.
- the purpose of error correction codes is to make digital communications more reliable by reducing the error rate.
- Error correcting coding systems introduce redundancy into the information to be transmitted using an encoder. At the reception level, the decoder uses this redundancy to try to find the original information, see Figure 02.
- the error correcting codes can be represented with a bipartite graph, see Figure 03. On one side there is the transmitted data and on the other side the test values. The decoder attempts to estimate the values of the data by exchanging trusts between the data and the test values. Confidence is calculated with an algorithm as the sum-product. These systems have very good performance, but they suffer from tremendous degradation if there are small cycles in the representative graph, see Figure 04. This problem has led to the use of large, hollow matrices for the coding of data. 'information. This use poses two problems, the first for the storage of the matrices and the second for the coding of the information.
- this decoder can generalize this decoder to IM levels by passing variables to their corresponding test values and then these to the interleaved variables of second level then to the test values then to the interleaved variables from the third level to the Nth levels to return to the initial variables see figure 06.
Landscapes
- Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Error Detection And Correction (AREA)
Abstract
The aim of the invention is to concatenate two codes in parallel following the interleaving of the data. The decoding is no longer carried out by the exchange of certainties between the nodes of variables and the corresponding test values thereof, but rather the certainties are transmitted from the variables to the corresponding test values thereof, then said test values to the interleaved variables, then from the interleaved variables to the corresponding test values thereof, and from the latter back to the variables (see figure 5). Said solution contributes to increasing the size of the cycles, and subsequently improving the performance of the system and even the use of small or medium-sized matrices. Said decoder can be generalised to N levels, from variables to the corresponding test values thereof, then the latter to the interleaved variables of the second level, then to the test values, then to the interleaved variables of the third level, up to the Nth level, to then return to the initial variables (see figure 6).
Description
DECODEUR PAR PROPAGATION DE CROYANCE GENERALISE A DES CODES AUTRE QUE LDPC GENERALIZED BELIEF PROPAGATION DECODER TO CODES OTHER THAN LDPC
Description Description
La présente invention concerne un décodeur pour code correcteur d'erreur qui étend les décodeurs à base de "BeliefPropagation" ou propagation de confiance. Ce décodeur minimise les limitations de ces derniers en ne se limitant plus à des matrices creuses pour le codage d'information. The present invention relates to a decoder for error correction code that extends the decoders based on "BeliefPropagation" or propagation of confidence. This decoder minimizes the limitations of the latter by no longer being limited to hollow matrices for the coding of information.
Décodeur a base de propagation de confiance étendu, permet d'améliorer la performance de décodage même pour les matrices de codage creuse de petite ou moyenne taille et même pour des matrice non creuse. Il permet une implémentation matérielle plus aisée que les décodeurs classiques comme les LDPC. Decoder based on extended confidence propagation, can improve the decoding performance even for small or medium size hollow coding matrices and even for non-hollow matrix. It allows easier hardware implementation than conventional decoders such as LDPCs.
Les communications numériques reposent sur un modèle de transmission composede trois parties. Lapremièrepartie est l'émetteur, la deuxième est le canal de transmission, et la troisièmepartie est le récepteur, voir Figure 01. Le canal de transmission est toujours sujet a des perturbations ou de bruits. Ces perturbations induisent des erreurs dans l'information a transmettre. Les codes correcteurs d'erreurs ont pour rôle de fiabiliser les communications numériques en réduisant le taux d'erreurs. Digital communications are based on a three-part transmission model. The first part is the transmitter, the second is the transmission channel, and the third part is the receiver, see Figure 01. The transmission channel is always subject to disturbances or noises. These disturbances induce errors in the information to be transmitted. The purpose of error correction codes is to make digital communications more reliable by reducing the error rate.
Les systèmes de codage correcteur d'erreurs introduisent une redondance dans l'information à transmettre en utilisant un codeur. Au niveau réception, le décodeur utilise cette redondance pour essayer de retrouver l'information d'origine, voir Figure 02. Error correcting coding systems introduce redundancy into the information to be transmitted using an encoder. At the reception level, the decoder uses this redundancy to try to find the original information, see Figure 02.
Parmi les systèmes de codage correcteur d'erreur on trouve ceux basés sur la propagation de confiance ou"BeliefPropagation" . Les codes correcteurs d'erreurs peuvent être représentés avec un graphe bipartite, voir Figure 03, d'un côté il y a les données transmises et de l'autre côté les valeurs de tests. Le décodeur essaye d'estimer les valeurs des données en échangeant les confiances entres les données et les valeurs de test. La confiance est calculée avec un algorithme comme la somme-produit. Cessystèmes ont de très bonnes performances, mais ils souffrent de dégradation énorme s'il y a des cycles de petite taille dans le graphe représentatif, voir Figure 04. Ce problème a poussé â utiliser des matrices creuses et de grandes tailles pour le codage de l'information. Cette utilisation pose deux problèmes, le premier pour le stockage des matrices et le deuxième pour le codage de l'information . Among the error-correcting coding systems are those based on trust propagation or "BeliefPropagation". The error correcting codes can be represented with a bipartite graph, see Figure 03. On one side there is the transmitted data and on the other side the test values. The decoder attempts to estimate the values of the data by exchanging trusts between the data and the test values. Confidence is calculated with an algorithm as the sum-product. These systems have very good performance, but they suffer from tremendous degradation if there are small cycles in the representative graph, see Figure 04. This problem has led to the use of large, hollow matrices for the coding of data. 'information. This use poses two problems, the first for the storage of the matrices and the second for the coding of the information.
Le décodage ne sera plus fait par l'échange des confiances entre les noeuds de variables et leurs valeurs de test correspondant, mais les confiance vont être transmis des variables vers leurs valeurs de test correspondant puis ces derniers valeurs de test vers les variables entrelacées puis des variables entrelacées vers leurs valeurs de test correspondants et depuis ces derniers en retourne vers les variables voir Figure 05. Cette solution va participer à augmenter la taille des cycles et par la suite améliorer les performances du système et même utiliser des matrices de moyenne ou petite taille. The decoding will no longer be done by the exchange of confidences between the variable nodes and their corresponding test values, but the confidence will be transmitted from the variables to their corresponding test values and then these last test values to the interleaved variables then the interlaced variables to their corresponding test values and from there to the variables see Figure 05. This solution will help to increase the size of the cycles and subsequently improve the performance of the system and even use matrices of medium or small size .
En peut généraliser ce décodeur é IM niveaux en passant des variables vers leurs valeurs de test correspondants puis ces derniers vers les variables entrelacées du
deuxième niveaux puis vers les valeurs de test puis vers les variables entrelacées du troisième niveau jusqu'au Nième Niveaux pour revenir aux variables initiales voir figure 06. It can generalize this decoder to IM levels by passing variables to their corresponding test values and then these to the interleaved variables of second level then to the test values then to the interleaved variables from the third level to the Nth levels to return to the initial variables see figure 06.
Figure 01 : modèle de communication numérique Figure 01: Digital Communication Model
Figure 02 : Fiabilisation des communications numériques Figure 02: Reliability of digital communications
Figure 03 : Décodeur a base de propagation de confiance Figure 03: Decoder based on confidence propagation
Figure 04 : Présence de cycle dans le graphe bipartite Figure 04: Presence of cycle in the bipartite graph
Figure 05 : Décodeur étendu à deux niveaux Figure 05: Extended two-level decoder
Figure 06 : Décodeur étendu à N niveaux
Figure 06: N-level extended decoder
Claims
1. Décodeur à base de propagation de confiance par concaténation parallèle 1. Decoder based on confidence propagation by parallel concatenation
Comprenant Comprising
deux concaténers two concateners
des matrices creuses hollow dies
deux codes en parallèle oùles variables sont entrelacées pour chaque niveau. Les confiances passent d'un niveau à l'autre pour revenir à la fin au premier niveau. two codes in parallel where the variables are interleaved for each level. The trusts move from one level to the next to return to the end at the first level.
Caractérisé en ce que l'effet fiabilité des communications numériques peut être obtenu par des codes correcteurs d'erreurs et la confiance est calculée avec un algorithme. Characterized in that the reliability effect of the digital communications can be obtained by error correcting codes and the confidence is calculated with an algorithm.
2. Décodeur é base de propagation de confiance caractérisé selon la revendication 1 en ce que l'utilisation et l'extension des cycles par la concaténation parallèle et les connections cascadées généralisent ce décodeur à N niveaux en passant des variables vers leurs valeurs de test correspondants puis ces derniers vers les variables entrelacées du deuxième niveaux puis vers les valeurs de test puis vers les variables entrelacées du troisième niveau jusqu'au Nième Niveaux pour revenir aux variables initiales
2. Confidence propagation base decoder characterized according to claim 1 in that the use and extension of the cycles by parallel concatenation and cascaded connections generalize this N-level decoder by passing variables to their corresponding test values. then these to the interleaved variables of the second level then to the test values and then to the interleaved variables from the third level to the Nth Levels to return to the initial variables
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| MA36996 | 2014-05-08 | ||
| MA36996A MA36996B1 (en) | 2014-05-08 | 2014-05-08 | Generalized decoder with codes other than idpc, low density parity check |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2015170949A1 true WO2015170949A1 (en) | 2015-11-12 |
Family
ID=52282822
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/MA2014/000026 WO2015170949A1 (en) | 2014-05-08 | 2014-11-04 | Generalised decoder employing belief propagation using codes other than ldpc codes |
Country Status (2)
| Country | Link |
|---|---|
| MA (1) | MA36996B1 (en) |
| WO (1) | WO2015170949A1 (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1531552A1 (en) * | 2003-11-14 | 2005-05-18 | Samsung Electronics Co., Ltd. | Channel encoding/decoding apparatus and method using a parallel concatenated low density parity check code |
| US20050160351A1 (en) * | 2003-12-26 | 2005-07-21 | Ko Young J. | Method of forming parity check matrix for parallel concatenated LDPC code |
-
2014
- 2014-05-08 MA MA36996A patent/MA36996B1/en unknown
- 2014-11-04 WO PCT/MA2014/000026 patent/WO2015170949A1/en active Application Filing
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1531552A1 (en) * | 2003-11-14 | 2005-05-18 | Samsung Electronics Co., Ltd. | Channel encoding/decoding apparatus and method using a parallel concatenated low density parity check code |
| US20050160351A1 (en) * | 2003-12-26 | 2005-07-21 | Ko Young J. | Method of forming parity check matrix for parallel concatenated LDPC code |
Non-Patent Citations (1)
| Title |
|---|
| MESHKAT PEYMAN ET AL: "Generalized versions of turbo decoding in the framework of Bayesian networks and Pearl's belief propagation algorithm", 1998 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS, ICC 98, ATLANTA, GA, USA 7-11 JUNE 1998, NEW YORK, NY, USA,IEEE, US, vol. 1, 7 June 1998 (1998-06-07), pages 121 - 125, XP010284547, ISBN: 978-0-7803-4788-5, DOI: 10.1109/ICC.1998.682598 * |
Also Published As
| Publication number | Publication date |
|---|---|
| MA36996B1 (en) | 2017-05-31 |
| MA36996A1 (en) | 2016-09-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105763203B (en) | Multi-element LDPC code decoding method based on hard reliability information | |
| US8245116B2 (en) | Method for performing soft decision decoding of Euclidean space Reed-Muller codes | |
| KR101751497B1 (en) | Apparatus and method using matrix network coding | |
| US9960790B2 (en) | Belief propagation decoding for short algebraic codes with permutations within the code space | |
| CN106059712B (en) | High-error-code arbitrary-code-rate convolutional code coding parameter blind identification method | |
| US11082147B2 (en) | Processing method, device and system for overlap multiplexing system | |
| US20150106680A1 (en) | Multiple component codes based generalized low-density parity-check codes for high-speed optical transport | |
| KR102208630B1 (en) | Method and system for estimating parameter of data channel model in a communication system | |
| Krainyk et al. | Low-complexity high-speed soft-hard decoding for turbo-product codes | |
| US20090019334A1 (en) | Error correction system using concatenated codes | |
| US8386877B2 (en) | Communication system, transmitter, error correcting code retransmitting method, and communication program | |
| US8413025B2 (en) | Method of handling packet loss using error-correcting codes and block rearrangement | |
| Arshad et al. | Implementation and analysis of convolutional codes using MATLAB | |
| US8627187B2 (en) | Decoding of recursive convolutional codes by means of a decoder for non-recursive convolutional codes | |
| US20070250760A1 (en) | Extended Convolutional Codes | |
| Lu et al. | Blind identification of convolutional interleaver parameters | |
| WO2015170949A1 (en) | Generalised decoder employing belief propagation using codes other than ldpc codes | |
| RU2637487C1 (en) | Method of decoding information using convolutional codes | |
| Chen et al. | Low-density lattice coded relaying with joint iterative decoding | |
| WO2022199529A1 (en) | Data coding processing method and apparatus, and storage medium and electronic apparatus | |
| JP5523064B2 (en) | Decoding apparatus and method | |
| RU2734450C2 (en) | Method for decoding of noise-immune codes | |
| Liu et al. | Damage‐resistance matrix embedding framework: the contradiction between robustness and embedding efficiency | |
| US20080222498A1 (en) | Sequential decoding method and apparatus thereof | |
| RU2667370C1 (en) | Method for decoding linear cascade code |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14824144 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 14824144 Country of ref document: EP Kind code of ref document: A1 |