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

Federated Transfer Learning for Two-Stage Decoding of LDPC Codes in 6G Networks

Quoc Bao Phan, Linh Nguyen, Tuy Tan Nguyen

Published Nov 12, 2024• 0 citations
Publisher

Abstract

The fifth-generation new radio coding techniques, such as low-density parity-check (LDPC) codes, have made significant progress. However, their limitations in the low error floor region prevent them from meeting the stringent bit error rate (BER) targets required for sixth-generation (6G) communications. 6G aims to support services with BER as low as 10 −9 , providing the reliability needed for mission-critical applications. his is particularly crucial for ultra-reliable lowlatency communication. In this paper, we present a two-stage decoding method that combines the strengths of layered LDPC decoding and federated transfer learning (FTL). The layered LDPC decoding stage effectively handles the challenging waterfall region, reducing errors during the initial decoding process. In addition, the FTL model adapts to mitigate the error floor and accelerate its decline. This combination achieves lower error rates and faster performance improvements, reducing the need for retransmissions and enhancing overall system efficiency.

DOI
10.36227/techrxiv.173145073.30283136/v1

Related papers

Federated Transfer Learning for Two-Stage Decoding of LDPC Codes

Linh Nguyen, Quoc Bao Phan, Cheol-Hong Min +1 more

The performance of fifth-generation (5G) coding techniques, such as low-density parity-check (LDPC) codes, is significantly affected by the error floor, which occurs in high signal-to-noise ratio (SNR) regions. Federated transfer learning (FTL), with its ability to learn from distributed data, can effectively improve error correction in low-SNR environments and extract crucial decoding features for high-SNR conditions. This paper introduces a two-stage decoding method that integrates layered LDPC decoding with FTL to mitigate the error floor and enhance system performance. Experimental results demonstrate that this approach significantly reduces error rates and improves decoding efficiency.

IEEE International Midwest Symposium on Circuits and Systems (MWSCAS)•Aug 10, 2025•0 citations

Boosted Neural Decoders: Achieving Extreme Reliability of LDPC Codes for 6G Networks

Hee-Youl Kwak, Dae-Young Yun, Yongjune Kim +2 more

Ensuring extremely high reliability in channel coding is essential for 6G networks. The next-generation of ultra-reliable and low-latency communications (xURLLC) scenario within 6G networks requires frame error rate (FER) below 10-9. However, low-density parity-check (LDPC) codes, the standard in 5G new radio (NR), encounter a challenge known as the error floor phenomenon, which hinders to achieve such low frame error rates. To tackle this problem, we introduce an innovative solution: boosted neural min-sum (NMS) decoder. This decoder operates identically to conventional NMS decoders, but is trained by novel training methods including: i) boosting learning with uncorrected vectors, ii) block-wise training schedule to address the vanishing gradient issue, iii) dynamic weight sharing to minimize the number of trainable weights, iv) transfer learning to reduce the required sample count, and v) data augmentation to expedite the sampling process. Leveraging these training strategies, the boosted NMS decoder achieves the state-of-the art performance in reducing the error floor as well as superior waterfall performance. Remarkably, we fulfill the 6G xURLLC requirement for 5G LDPC codes without a severe error floor. Additionally, the boosted NMS decoder, once its weights are trained, can perform decoding without additional modules, making it highly practical for immediate application. The source code is available athttps://github.com/ghy1228/LDPC_Error_Floor.

arXiv:2405.13413•IEEE Journal on Selected Areas in Communications•Jan 20, 2025•18 citations

Performance Simulation of a Layered QC-LDPC Decoder for URLLC Applications

Bilal Mejmaa, Chakir Aqil, Ismail Akharraz +1 more

This research investigates the simulation performance of a layered Quasi-Cyclic Low-Density Parity-Check (QC-LDPC) decoder intended for Ultra-Reliable Low-Latency Communication (URLLC) applications in 5G New-Radio (NR) standards. The proposed work focuses on short block-length QC-LDPC codes (280,176) to meet the dual challenge of achieving ultra-low latency (<1ms) while maintaining high reliability. LDPC codes have traditionally excelled at longer code lengths; this study demonstrates the effectiveness of LDPC codes in short-length contexts through architectural optimization utilizing a layered decoding approach, which provides improved convergence speed and reduced latency compared to traditional flooding techniques. The baseline QC-LDPC decoder performs better in terms of Bit Error Rate (BER) and Block Error Rate (BLER) with only 7 iterations. For a target BLER of 10 −2 , the decoder outperforms reference shorter codes (K40, K64) with an E b /N o gain of up to 0.9 dB.

EPJ Web of Conferences•Jan 1, 2026•0 citations

Performance Simulation of a Layered QC-LDPC Decoder for URLLC Applications

Bilal Mejmaa, Chakir Aqil, Ismail Akharraz +1 more

This research investigates the simulation performance of a layered Quasi-Cyclic Low-Density Parity-Check (QC-LDPC) decoder intended for Ultra-Reliable Low-Latency Communication (URLLC) applications in 5G New-Radio (NR) standards. The proposed work focuses on short block-length QC-LDPC codes (280,176) to meet the dual challenge of achieving ultra-low latency (<1ms) while maintaining high reliability. LDPC codes have traditionally excelled at longer code lengths; this study demonstrates the effectiveness of LDPC codes in short-length contexts through architectural optimization utilizing a layered decoding approach, which provides improved convergence speed and reduced latency compared to traditional flooding techniques. The baseline QC-LDPC decoder performs better in terms of Bit Error Rate (BER) and Block Error Rate (BLER) with only 7 iterations. For a target BLER of 10−2, the decoder outperforms reference shorter codes (K40, K64) with an Eb/No gain of up to 0.9 dB.

•Jul 1, 2026•0 citations

Intelligent DRL-assisted decoding of error-correcting codes for 5G/6G telecommunication channels

Juliy Boiko, Ilya Pyatin, Oleksander Eromenko

Abstract This paper presents an intelligent decoding methodology for low-density parity-check (LDPC) and Polar codes (P-C) based on deep reinforcement learning (DRL+Soft) for modern 5G and emerging 6G communication systems. The proposed decoder adapts its decoding strategy in real-time, optimizing bit error rate (BER) performance under varying channel conditions. Evaluation is formulated through performance metrics including BER convergence, reward dynamics, computational complexity (FLOPs), and inference latency. Simulation results show that the DRL+Soft decoder achieves up to a 2.5-fold reduction in computational cost and decreases inference time from 4.8 ms (BP) to 1.6 ms per data block, while maintaining superior BER compared to classical belief propagation (BP) and cyclic redundancy check-aided successive cancellation list (CA-SCL) decoders. The decoder exhibits improved robustness against unpredictable interference and channel impairments, making it suitable for dense device deployments and ultra-reliable low-latency communication (URLLC) scenarios. The methodology also supports reprogramming or retraining of DRL agents without hardware changes, ensuring long-term adaptability for evolving 6G networks. These results demonstrate both theoretical novelty and practical value, providing a scalable and energy-efficient solution for future wireless communication systems.

Journal of Electrical Engineering•Dec 1, 2025•4 citations

Hybrid Optimization Framework for Enhancing LDPC Base Graph Performance in 6G Networks: Integrating Machine Learning and Genetic Algorithms

This study presents a novel hybrid framework designed to enhance Low-Density Parity-Check (LDPC) code performance for 6G networks by integrating multi-stage optimization and adaptive neural decoding.First, density evolution simulations quantify bit error rate (BER) and frame error rate (FER) across Additive White Gaussian Noise (AWGN), Binary Symmetric Channel (BSC), and Rayleigh flat-fading channel.Second, a genetic algorithm (GA), guided by a LightGBM (ML) predictor, tunes Base Graph (BG) shift values to reduce error rates and short cycles in the parity check matrix.Third, a custom Neural Min-Sum decoder incorporates learnable weights and syndrome-aware loss, leveraging uncorrected frame samples to expedite convergence and achieve ultra-low error floors.The simulation results indicate that the optimized design improved LDPC BGs by up to 44% for the AWGN channel, 94% for the BSC, and 61.3% for the Rayleigh channel, leading to enhanced BER performance compared to the original designs.Additionally, the computation time is reduced by approximately 85%, and the number of short cycles (4-cycles) is decreased by about 34% using QPSK modulation.

International journal of intelligent engineering and systems•Jul 18, 2025•0 citations
The fifth-generation new radio coding techniques, such as low-density parity-check (LDPC) codes, have made significant progress. However, their limitations in the low error floor region prevent them from meeting the stringent bit error rate (BER) targets required for sixth-generation (6G) communications. 6G aims to support services with BER as low as 10 −9 , providing the reliability needed for mission-critical applications. his is particularly crucial for ultra-reliable lowlatency communication. In this paper, we present a two-stage decoding method that combines the strengths of layered LDPC decoding and federated transfer learning (FTL). The layered LDPC decoding stage effectively handles the challenging waterfall region, reducing errors during the initial decoding process. In addition, the FTL model adapts to mitigate the error floor and accelerate its decline. This combination achieves lower error rates and faster performance improvements, reducing the need for retransmissions and enhancing overall system efficiency.
The performance of fifth-generation (5G) coding techniques, such as low-density parity-check (LDPC) codes, is significantly affected by the error floor, which occurs in high signal-to-noise ratio (SNR) regions. Federated transfer learning (FTL), with its ability to learn from distributed data, can effectively improve error correction in low-SNR environments and extract crucial decoding features for high-SNR conditions. This paper introduces a two-stage decoding method that integrates layered LDPC decoding with FTL to mitigate the error floor and enhance system performance. Experimental results demonstrate that this approach significantly reduces error rates and improves decoding efficiency.
Ensuring extremely high reliability in channel coding is essential for 6G networks. The next-generation of ultra-reliable and low-latency communications (xURLLC) scenario within 6G networks requires frame error rate (FER) below 10-9. However, low-density parity-check (LDPC) codes, the standard in 5G new radio (NR), encounter a challenge known as the error floor phenomenon, which hinders to achieve such low frame error rates. To tackle this problem, we introduce an innovative solution: boosted neural min-sum (NMS) decoder. This decoder operates identically to conventional NMS decoders, but is trained by novel training methods including: i) boosting learning with uncorrected vectors, ii) block-wise training schedule to address the vanishing gradient issue, iii) dynamic weight sharing to minimize the number of trainable weights, iv) transfer learning to reduce the required sample count, and v) data augmentation to expedite the sampling process. Leveraging these training strategies, the boosted NMS decoder achieves the state-of-the art performance in reducing the error floor as well as superior waterfall performance. Remarkably, we fulfill the 6G xURLLC requirement for 5G LDPC codes without a severe error floor. Additionally, the boosted NMS decoder, once its weights are trained, can perform decoding without additional modules, making it highly practical for immediate application. The source code is available athttps://github.com/ghy1228/LDPC_Error_Floor.
This research investigates the simulation performance of a layered Quasi-Cyclic Low-Density Parity-Check (QC-LDPC) decoder intended for Ultra-Reliable Low-Latency Communication (URLLC) applications in 5G New-Radio (NR) standards. The proposed work focuses on short block-length QC-LDPC codes (280,176) to meet the dual challenge of achieving ultra-low latency (<1ms) while maintaining high reliability. LDPC codes have traditionally excelled at longer code lengths; this study demonstrates the effectiveness of LDPC codes in short-length contexts through architectural optimization utilizing a layered decoding approach, which provides improved convergence speed and reduced latency compared to traditional flooding techniques. The baseline QC-LDPC decoder performs better in terms of Bit Error Rate (BER) and Block Error Rate (BLER) with only 7 iterations. For a target BLER of 10 −2 , the decoder outperforms reference shorter codes (K40, K64) with an E b /N o gain of up to 0.9 dB.
This research investigates the simulation performance of a layered Quasi-Cyclic Low-Density Parity-Check (QC-LDPC) decoder intended for Ultra-Reliable Low-Latency Communication (URLLC) applications in 5G New-Radio (NR) standards. The proposed work focuses on short block-length QC-LDPC codes (280,176) to meet the dual challenge of achieving ultra-low latency (<1ms) while maintaining high reliability. LDPC codes have traditionally excelled at longer code lengths; this study demonstrates the effectiveness of LDPC codes in short-length contexts through architectural optimization utilizing a layered decoding approach, which provides improved convergence speed and reduced latency compared to traditional flooding techniques. The baseline QC-LDPC decoder performs better in terms of Bit Error Rate (BER) and Block Error Rate (BLER) with only 7 iterations. For a target BLER of 10−2, the decoder outperforms reference shorter codes (K40, K64) with an Eb/No gain of up to 0.9 dB.
Abstract This paper presents an intelligent decoding methodology for low-density parity-check (LDPC) and Polar codes (P-C) based on deep reinforcement learning (DRL+Soft) for modern 5G and emerging 6G communication systems. The proposed decoder adapts its decoding strategy in real-time, optimizing bit error rate (BER) performance under varying channel conditions. Evaluation is formulated through performance metrics including BER convergence, reward dynamics, computational complexity (FLOPs), and inference latency. Simulation results show that the DRL+Soft decoder achieves up to a 2.5-fold reduction in computational cost and decreases inference time from 4.8 ms (BP) to 1.6 ms per data block, while maintaining superior BER compared to classical belief propagation (BP) and cyclic redundancy check-aided successive cancellation list (CA-SCL) decoders. The decoder exhibits improved robustness against unpredictable interference and channel impairments, making it suitable for dense device deployments and ultra-reliable low-latency communication (URLLC) scenarios. The methodology also supports reprogramming or retraining of DRL agents without hardware changes, ensuring long-term adaptability for evolving 6G networks. These results demonstrate both theoretical novelty and practical value, providing a scalable and energy-efficient solution for future wireless communication systems.
This study presents a novel hybrid framework designed to enhance Low-Density Parity-Check (LDPC) code performance for 6G networks by integrating multi-stage optimization and adaptive neural decoding.First, density evolution simulations quantify bit error rate (BER) and frame error rate (FER) across Additive White Gaussian Noise (AWGN), Binary Symmetric Channel (BSC), and Rayleigh flat-fading channel.Second, a genetic algorithm (GA), guided by a LightGBM (ML) predictor, tunes Base Graph (BG) shift values to reduce error rates and short cycles in the parity check matrix.Third, a custom Neural Min-Sum decoder incorporates learnable weights and syndrome-aware loss, leveraging uncorrected frame samples to expedite convergence and achieve ultra-low error floors.The simulation results indicate that the optimized design improved LDPC BGs by up to 44% for the AWGN channel, 94% for the BSC, and 61.3% for the Rayleigh channel, leading to enhanced BER performance compared to the original designs.Additionally, the computation time is reduced by approximately 85%, and the number of short cycles (4-cycles) is decreased by about 34% using QPSK modulation.