Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

IEEE International Symposium on Information Theory (ISIT)

Deep Learning-Based Bit Reliability Based Decoding for Non-binary LDPC Codes

Taishi Watanabe, Takeo Ohseki, Kosuke Yamazaki

Published Jul 12, 2021• 6 citations
Publisher

Abstract

The bit reliability based (BRB) and weighted bit reliability based (wBRB) algorithms are non-binary low-density parity-check (LDPC) code decoding algorithms with an excellent tradeoff between computational complexity and performance. However, the performance of these algorithms needs further improvement. We apply deep learning to these algorithms. Weights are assigned to each edge of the Tanner graphs of the non-binary LDPC codes in the proposed algorithms. We demonstrate the effectiveness of applying deep learning to the BRB and wBRB algorithms in terms of implementation and performance. The proposed algorithms achieve an approximately 0.3 dB higher bit error rate performance than the original algorithms in the high SNR region. The increase in computational complexity and memory consumption does not significantly change the implementation of the algorithms.

DOI
10.1109/isit45174.2021.9517790

Related papers

Bit Reliability-Based Decoders for Non-Binary LDPC Codes

Qin Huang, Shuai Yuan

Message-passing decoders typically perform well for nonbinary low-density parity-check (NB-LDPC) codes with large computational complexity. As another type of simplified decoders, symbol-reliability-based decoders further reduce the computational complexity. However, the previously proposed algorithms suffer severe error performance degradation for NB-LDPC codes with low column weights. In this paper, a weighted bit-reliability based (wBRB) decoder for NB-LDPC codes is developed and implemented with efficient layered partial-parallel structure. It not only balances the tradeoff between complexity and error performance, but also reduces the memory usage significantly. Furthermore, to enhance the performance of the wBRB decoder, a full bit-reliability-based (FBRB) decoder is proposed. The FBRB decoder is derived based on the binary matrix representation of the nonzero entries in the parity-check matrix. Since more bit-reliability values are passed through the edges of the Tanner graph, the FBRB decoder can achieve better error performance and faster convergence rate than the wBRB decoder. Both of the decoders are implemented on a Xilinx Virtex-5 XC5VLX155T FPGA device for a (403,226) code over GF(25). The results shows that they achieve 118.98 and 95.73 Mbps throughput with 15 iterations, respectively.

IEEE Transactions on Communications•Nov 17, 2015•26 citations

Low error-floor majority-logic decoding based algorithm for non-binary LDPC codes

Liyuan Song, Mu Zhang, Qin Huang +1 more

The traditional majority-logic decoding (MLgD) based algorithms suffer error-floors for decoding non-binary LDPC codes with small column weights. This paper presents a bit-reliability based MLgD (BRB-MLgD) algorithm with low error-floors for non-binary LDPC codes. The proposed algorithm is carried out based on the binary representations of non-binary symbols. The reliability update along each edge of the Tanner graph of a non-binary LDPC code is in terms of bits rather than symbols. Thus, its computational complexity and memory consumption are less than those of the existing MLgD based algorithms. Simulation results indicate that the proposed algorithm can significantly reduce error-floors with small performance degradation in the waterfall region.

International Conference on Communications•Jun 1, 2015•2 citations

Hard-Information Bit-Reliability Based Decoding Algorithm for Majority-Logic Decodable Nonbinary LDPC Codes

Xiangcheng Li, Tuanfa Qin, Haiqiang Chen +3 more

A modified bit-reliability based decoding algorithm is presented based on a recent work by Huang et al. For the presented algorithm, only one Galois field element is passed and exchanged along the edges of the Tanner graph. At variable nodes, full messages rather than extrinsic messages are processed to further reduce the computational complexity. At check nodes, only hard reliability is considered and the main operation is to compute the check-sum and send one extrinsic symbol back to variable nodes. Simulation results show that, even with lower complexity and less memory consumption, the presented algorithm still can perform very closely to the original wBRB algorithm with low quantization bits (3 ~ 4 bits) when decoding the majority-logic decodable nonbinary LDPC codes. For codes constructed in high order fields, the presented algorithm can even outperform the original wBRB algorithm.

IEEE Communications Letters•Mar 3, 2016•10 citations

Bit-Reliability Based Low-Complexity Decoding Algorithms for Non-Binary LDPC Codes

Qin Huang, Mu Zhang, Zulin Wang +1 more

This paper presents bit-reliability based majority-logic decoding (MLgD) algorithms for non-binary LDPC codes. The proposed algorithms pass only one Galois field element and its reliability along each edge of the Tanner graph of a non-binary LDPC code. Since their reliability updates are in terms of bits rather than symbols, they are more efficient than traditional MLgD based decoding algorithms. By weighting the soft reliability of the extrinsic information-sums based on their hard reliability, the proposed algorithms can achieve good error performance for non-binary LDPC codes with various column weights. Moreover, their computational complexity and memory consumption are remarkably reduced compared with existing MLgD based decoding algorithms. As a result, they provide effective tradeoffs between error performance and complexity for decoding of non-binary LDPC codes.

IEEE Transactions on Communications•Nov 12, 2014•36 citations

A BRB Algorithm for Non-Binary LDPC Decoding With Single-Symbol Extrinsic Information Modification

Xiangcheng Li, Ziyun Ruan, Dongri Ban +2 more

In the weighted bit-reliability based (wBRB) Non-binary LDPC decoding algorithm, using the minimum bit-reliability as the reliability of the entire symbol for information updating significantly decreases the decoding performance in cases of a large order of Galois field or small column weight. The improved algorithm——MwBRB increases the decoding performance by introducing three kinds of extrinsic information transfer but suffers from high decoding complexity. To address these problems, a BRB Non-binary LDPC decoding algorithm based on single-symbol extrinsic information modification is proposed. At the variable node$V_{j}$, using a symbol-related reliability calculation method, the maximum reliability information and its corresponding most possible Galois field symbol are obtained, and then transferred to the check node$C_{i}$; at the check node$C_{i}$, only the most possible Galois field symbol is used for extrinsic information processing, and its reliability of the extrinsic information is represented by the minimum reliability of other Galois field symbols. Finally, at the check node$C_{i}$, the extrinsic information is processed by based on the Hamming distance coefficient in terms of bits. Simulation and performance analysis show that the proposed algorithm reduces the complexity by over 50% compared to the MwBRB algorithm, and maintains consistent or slightly superior decoding performance. Additionally, the proposed algorithm exhibits similar decoding performance to LDPC codes constructed by different methods and shows better robustness.

IEEE Access•Jan 1, 2024•0 citations

E-IHRB Decoder Architectures for Non-Binary LDPC Codes

Madamsetti, Kalyani, M, Suseela

In the information theory , the non- binary low density parity check  (NB-LDPC) code is a  linear error correcting code , a method of transmitting message over a  noisy transmission channel. LDPCcodes are  capacity-approaching codes , which  means  that  practical constructions exist  that allow the noise threshold to be set very close to the theoretical maximum for a symmetric memory less channel.Using iterative  belief propagation techniques, NB-LDPC codes can be decoded in time linear to their block length. Non-binary low-density parity-check codes can achieve higher performance than their binary counterparts. Recently, two algorithms were developed for decoding NB-LDPC codes: iterative hard reliability-based majority-logic decoding (IHRB-MLGD) and iterative soft reliability-based majority-logic decoding (ISRB-MLGD has better performance-complexity tradeoffs than previous algorithms. This paper first proposes message-passing in such a way that we can reduce the memory consumption to a greater extent at the same time there should not be performance loss. By using enhanced iterative reliability-based majority-logic decoding algorithms we can achieve effective complexity-performance and noise free design. Compared to previous designs, our proposed decoders have at least tens of times lower complexity with moderate coding gain loss.

•Aug 31, 2014•0 citations
The bit reliability based (BRB) and weighted bit reliability based (wBRB) algorithms are non-binary low-density parity-check (LDPC) code decoding algorithms with an excellent tradeoff between computational complexity and performance. However, the performance of these algorithms needs further improvement. We apply deep learning to these algorithms. Weights are assigned to each edge of the Tanner graphs of the non-binary LDPC codes in the proposed algorithms. We demonstrate the effectiveness of applying deep learning to the BRB and wBRB algorithms in terms of implementation and performance. The proposed algorithms achieve an approximately 0.3 dB higher bit error rate performance than the original algorithms in the high SNR region. The increase in computational complexity and memory consumption does not significantly change the implementation of the algorithms.
Message-passing decoders typically perform well for nonbinary low-density parity-check (NB-LDPC) codes with large computational complexity. As another type of simplified decoders, symbol-reliability-based decoders further reduce the computational complexity. However, the previously proposed algorithms suffer severe error performance degradation for NB-LDPC codes with low column weights. In this paper, a weighted bit-reliability based (wBRB) decoder for NB-LDPC codes is developed and implemented with efficient layered partial-parallel structure. It not only balances the tradeoff between complexity and error performance, but also reduces the memory usage significantly. Furthermore, to enhance the performance of the wBRB decoder, a full bit-reliability-based (FBRB) decoder is proposed. The FBRB decoder is derived based on the binary matrix representation of the nonzero entries in the parity-check matrix. Since more bit-reliability values are passed through the edges of the Tanner graph, the FBRB decoder can achieve better error performance and faster convergence rate than the wBRB decoder. Both of the decoders are implemented on a Xilinx Virtex-5 XC5VLX155T FPGA device for a (403,226) code over GF(25). The results shows that they achieve 118.98 and 95.73 Mbps throughput with 15 iterations, respectively.
The traditional majority-logic decoding (MLgD) based algorithms suffer error-floors for decoding non-binary LDPC codes with small column weights. This paper presents a bit-reliability based MLgD (BRB-MLgD) algorithm with low error-floors for non-binary LDPC codes. The proposed algorithm is carried out based on the binary representations of non-binary symbols. The reliability update along each edge of the Tanner graph of a non-binary LDPC code is in terms of bits rather than symbols. Thus, its computational complexity and memory consumption are less than those of the existing MLgD based algorithms. Simulation results indicate that the proposed algorithm can significantly reduce error-floors with small performance degradation in the waterfall region.
A modified bit-reliability based decoding algorithm is presented based on a recent work by Huang et al. For the presented algorithm, only one Galois field element is passed and exchanged along the edges of the Tanner graph. At variable nodes, full messages rather than extrinsic messages are processed to further reduce the computational complexity. At check nodes, only hard reliability is considered and the main operation is to compute the check-sum and send one extrinsic symbol back to variable nodes. Simulation results show that, even with lower complexity and less memory consumption, the presented algorithm still can perform very closely to the original wBRB algorithm with low quantization bits (3 ~ 4 bits) when decoding the majority-logic decodable nonbinary LDPC codes. For codes constructed in high order fields, the presented algorithm can even outperform the original wBRB algorithm.
This paper presents bit-reliability based majority-logic decoding (MLgD) algorithms for non-binary LDPC codes. The proposed algorithms pass only one Galois field element and its reliability along each edge of the Tanner graph of a non-binary LDPC code. Since their reliability updates are in terms of bits rather than symbols, they are more efficient than traditional MLgD based decoding algorithms. By weighting the soft reliability of the extrinsic information-sums based on their hard reliability, the proposed algorithms can achieve good error performance for non-binary LDPC codes with various column weights. Moreover, their computational complexity and memory consumption are remarkably reduced compared with existing MLgD based decoding algorithms. As a result, they provide effective tradeoffs between error performance and complexity for decoding of non-binary LDPC codes.
In the weighted bit-reliability based (wBRB) Non-binary LDPC decoding algorithm, using the minimum bit-reliability as the reliability of the entire symbol for information updating significantly decreases the decoding performance in cases of a large order of Galois field or small column weight. The improved algorithm——MwBRB increases the decoding performance by introducing three kinds of extrinsic information transfer but suffers from high decoding complexity. To address these problems, a BRB Non-binary LDPC decoding algorithm based on single-symbol extrinsic information modification is proposed. At the variable nodeVjV_{j}Vj​, using a symbol-related reliability calculation method, the maximum reliability information and its corresponding most possible Galois field symbol are obtained, and then transferred to the check nodeCiC_{i}Ci​; at the check nodeCiC_{i}Ci​, only the most possible Galois field symbol is used for extrinsic information processing, and its reliability of the extrinsic information is represented by the minimum reliability of other Galois field symbols. Finally, at the check nodeCiC_{i}Ci​, the extrinsic information is processed by based on the Hamming distance coefficient in terms of bits. Simulation and performance analysis show that the proposed algorithm reduces the complexity by over 50% compared to the MwBRB algorithm, and maintains consistent or slightly superior decoding performance. Additionally, the proposed algorithm exhibits similar decoding performance to LDPC codes constructed by different methods and shows better robustness.
In the information theory , the non- binary low density parity check  (NB-LDPC) code is a  linear error correcting code , a method of transmitting message over a  noisy transmission channel. LDPCcodes are  capacity-approaching codes , which  means  that  practical constructions exist  that allow the noise threshold to be set very close to the theoretical maximum for a symmetric memory less channel.Using iterative  belief propagation techniques, NB-LDPC codes can be decoded in time linear to their block length. Non-binary low-density parity-check codes can achieve higher performance than their binary counterparts. Recently, two algorithms were developed for decoding NB-LDPC codes: iterative hard reliability-based majority-logic decoding (IHRB-MLGD) and iterative soft reliability-based majority-logic decoding (ISRB-MLGD has better performance-complexity tradeoffs than previous algorithms. This paper first proposes message-passing in such a way that we can reduce the memory consumption to a greater extent at the same time there should not be performance loss. By using enhanced iterative reliability-based majority-logic decoding algorithms we can achieve effective complexity-performance and noise free design. Compared to previous designs, our proposed decoders have at least tens of times lower complexity with moderate coding gain loss.