Hyper-Graph-Network Decoders for Block Codes
Eliya Nachmani, Lior Wolf

TL;DR
This paper introduces a novel hyper-graph-network decoder for algebraic block codes that outperforms traditional belief propagation and existing learning-based methods across various code families.
Contribution
It extends neural decoding to larger algebraic codes using hypernetworks for message passing, improving decoding performance and stability.
Findings
Outperforms belief propagation on BCH, LDPC, and Polar codes
Uses hypernetworks for adaptive message passing
Achieves better decoding accuracy than existing neural methods
Abstract
Neural decoders were shown to outperform classical message passing techniques for short BCH codes. In this work, we extend these results to much larger families of algebraic block codes, by performing message passing with graph neural networks. The parameters of the sub-network at each variable-node in the Tanner graph are obtained from a hypernetwork that receives the absolute values of the current message as input. To add stability, we employ a simplified version of the arctanh activation that is based on a high order Taylor approximation of this activation function. Our results show that for a large number of algebraic block codes, from diverse families of codes (BCH, LDPC, Polar), the decoding obtained with our method outperforms the vanilla belief propagation method as well as other learning techniques from the literature.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsError Correcting Code Techniques · DNA and Biological Computing · Gene expression and cancer classification
MethodsHyperNetwork
