RELDEC: Reinforcement Learning-Based Decoding of Moderate Length LDPC Codes

被引:3
|
作者
Habib, Salman [1 ]
Beemer, Allison [2 ]
Kliewer, Jorg [1 ]
机构
[1] New Jersey Inst Technol, Helen & John C Hartmann Dept Elect & Comp Engn, Newark, NJ 07102 USA
[2] Univ Wisconsin, Dept Math, Eau Claire, WI 54701 USA
关键词
Decoding; Schedules; Optimal scheduling; Iterative decoding; Metalearning; Signal to noise ratio; Clustering algorithms; Artificial intelligence; channel coding; reinforcement learning (RL); wireless communication; PARITY-CHECK CODES; DESIGN; 5G;
D O I
10.1109/TCOMM.2023.3296621
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
In this work we propose RELDEC, a novel approach for sequential decoding of moderate length low-density parity-check (LDPC) codes. The main idea behind RELDEC is that an optimized decoding policy is subsequently obtained via reinforcement learning based on a Markov decision process (MDP). In contrast to our previous work, where an agent learns to schedule only a single check node (CN) within a group (cluster) of CNs per iteration, in this work we train the agent to schedule all CNs in a cluster, and all clusters in every iteration. That is, in each learning step of RELDEC an agent learns to schedule CN clusters sequentially depending on a reward associated with the outcome of scheduling a particular cluster. We also modify the state space representation of the MDP, enabling RELDEC to be suitable for larger block length LDPC codes than those studied in our previous work. Furthermore, to address decoding under varying channel conditions, we propose agile meta-RELDEC (AM-RELDEC) that employs meta-reinforcement learning. The proposed RELDEC scheme significantly outperforms standard flooding and random sequential decoding for a variety of LDPC codes, including codes designed for 5G new radio.
引用
收藏
页码:5661 / 5674
页数:14
相关论文
共 50 条
  • [31] Dynamic Decoding Algorithms of LDPC Codes Based on Simulated Annealing
    Cui, Ru
    Liu, Xingcheng
    2014 INTERNATIONAL WORKSHOP ON HIGH MOBILITY WIRELESS COMMUNICATIONS (HMWC), 2014, : 135 - 139
  • [32] On Construction of Moderate-Length LDPC Codes over Correlated Erasure Channels
    Liva, Gianluigi
    Matuz, Balazs
    Katona, Zoltan
    Paolini, Enrico
    Chiani, Marco
    2009 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS, VOLS 1-8, 2009, : 3820 - +
  • [33] Analysis and design of moderate length regular LDPC codes with low error floors
    Cole, Chad A.
    Wilson, Stephen G.
    Hall, Eric. K.
    Giallorenzi, Thomas R.
    2006 40TH ANNUAL CONFERENCE ON INFORMATION SCIENCES AND SYSTEMS, VOLS 1-4, 2006, : 823 - 828
  • [34] Tree-Structured Expectation Propagation for Decoding Finite-Length LDPC Codes
    Olmos, Pablo M.
    Jose Murillo-Fuentes, Juan
    Perez-Cruz, Fernando
    IEEE COMMUNICATIONS LETTERS, 2011, 15 (02) : 235 - 237
  • [35] Approaching Maximum Likelihood decoding of finite length LDPC codes via FAID diversity
    Declercq, David
    Li, Erbao
    Vasic, Bane
    Pianjery, Shiva K.
    2012 IEEE INFORMATION THEORY WORKSHOP (ITW), 2012, : 487 - 491
  • [36] A decoding algorithm based on layered decoding and min-max for nonbinary LDPC codes
    Yang, W. (yangkk3@tju.edu.cn), 1677, Science Press (35):
  • [37] Reinforcement Learning for Bit-Flipping Decoding of Polar Codes
    Wang, Xiumin
    He, Jinlong
    Li, Jun
    Shan, Liang
    ENTROPY, 2021, 23 (02) : 1 - 16
  • [38] Deep Learning-Aided Trainable Projected Gradient Decoding for LDPC Codes
    Wadayama, Tadashi
    Takabe, Satoshi
    2019 IEEE INTERNATIONAL SYMPOSIUM ON INFORMATION THEORY (ISIT), 2019, : 2444 - 2448
  • [39] On a decoding algorithm for LDPC codes based on the concave-convex procedure
    Shibuya, T
    Sakaniwa, K
    2003 IEEE INTERNATIONAL SYMPOSIUM ON INFORMATION THEORY - PROCEEDINGS, 2003, : 148 - 148
  • [40] Enhanced verification-based decoding for packet-based LDPC codes
    Zhu, Bin
    Huang, Defeng David
    Nordholm, Sven
    IEEE COMMUNICATIONS LETTERS, 2008, 12 (02) : 136 - 138