Finite Alphabet Iterative Decoding of LDPC Codes with Coarsely Quantized Neural Networks
MetadataShow full item record
CitationX. Xiao, B. Vasic, R. Tandon and S. Lin, "Finite Alphabet Iterative Decoding of LDPC Codes with Coarsely Quantized Neural Networks," 2019 IEEE Global Communications Conference (GLOBECOM), Waikoloa, HI, USA, 2019, pp. 1-6, doi: 10.1109/GLOBECOM38437.2019.9013364.
RightsCopyright © 2019 IEEE.
Collection InformationThis item from the UA Faculty Publications collection is made available by the University of Arizona with support from the University of Arizona Libraries. If you have questions, please contact us at email@example.com.
AbstractIn this paper, we introduce a method of using quantized neural networks (QNN) to design finite alphabet message passing decoders (FAID) for Low-Density Parity Check (LDPC) codes. Specifically, we construct a neural network with low precision activations to optimize a FAID over Additive White Gaussian Noise Channel (AWGNC). The low precision activations cause a critical issue that their gradients vanish almost everywhere, making it difficult to use classical backward propagation. We introduce straight-through estimators (STE)  to avoid this problem, by replacing zero derivatives of quantized activations with surrogate gradients in the chain rules. We present a systematic approach to train such networks while minimizing the bit error rate, which is a widely used and accurate metric to measure the performance of iterative decoders. Examples and simulations show that by training a QNN, a FAID with 3-bit of message and 4-bit of channel output can be obtained, which performs better than the more complex floating-point minsum decoding algorithm. This methodology is promising in the sense that it facilitates designing low-precision FAID for LDPC codes while maintaining good error performance in a flexible and efficient manner.
VersionFinal accepted manuscript