Leveraging Code Automorphisms for Improved Syndrome-Based Neural Decoding

TL;DR

This paper enhances syndrome-based neural decoding by leveraging code automorphisms, improving model performance and generalization with data augmentation, and revealing undertraining in prior results.

Contribution

It introduces a method to incorporate code automorphisms into SBND, significantly boosting decoding accuracy and generalization, and provides insights into undertraining issues in previous studies.

Findings

Models closely approach MLD performance with small datasets.

Data augmentation via automorphisms improves learning and generalization.

Prior SBND results underestimated correction capabilities due to undertraining.

Abstract

Syndrome-based neural decoding (SBND) has emerged as a promising deep learning approach for soft-decision decoding of high-rate, short-length codes. However, this approach still has substantial room for improvement. In this paper, we show how to leverage code automorphisms to enhance the ability of existing SBND models to learn and generalize through data augmentation during training and inference. As a result, for the short high-rate codes considered, we obtain models that closely approach MLD performance using small datasets and proper training. Our findings also suggest that many prior results for SBND models in the literature underestimate their true correction capability due to undertraining. Code to reproduce all results is available at: https://github.com/lebidan/sbnd.