Learned decimation for neural belief propagation decoders (invited paper)

Conference Paper

We introduce a two-stage decimation process to improve the performance of neural belief propagation (NBP), recently introduced by Nachmani et al., for short low-density paritycheck (LDPC) codes. In the first stage, we build a list by iterating between a conventional NBP decoder and guessing the least reliable bit. The second stage iterates between a conventional NBP decoder and learned decimation, where we use a neural network to decide the decimation value for each bit. For a (128,64) LDPC code, the proposed NBP with decimation outperforms NBP decoding by 0.75 dB and performs within 1 dB from maximum-likelihood decoding at a block error rate of 10-4.

Full Text

Duke Authors

Cited Authors

  • Buchberger, A; Häger, C; Pfister, HD; Schmalen, L; I Amat, AG

Published Date

  • January 1, 2021

Published In

Volume / Issue

  • 2021-June /

Start / End Page

  • 8273 - 8277

International Standard Serial Number (ISSN)

  • 1520-6149

Digital Object Identifier (DOI)

  • 10.1109/ICASSP39728.2021.9414407

Citation Source

  • Scopus