Interleaver design and pairwise codeword distance distribution enhancement for turbo autoencoder

Patent metadata

Field	Value
Publication number	US-12175353-B2
Application number	US-202217726506-A
Country	US
Kind code	B2
Filing date	Apr 21, 2022
Priority date	May 21, 2021
Publication date	Dec 24, 2024
Grant date	Dec 24, 2024

Abstract

A symmetric interleaver for a Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) encoder and a circular padding mode are disclosed. The interleaver interleaves elements of an input block to form an output block in which an output neighborhood of elements for each element of the output block is symmetric to an input neighborhood of elements for each element of the input block. A position of an element of the input block is interleaved based on an index i of the position times a parameter δ modulo K in which the parameter δ is relatively prime with K. A test loss function may be used to train the encoder that includes a Binary Cross Entropy (BCE) loss function plus a function that minimizes a number of codeword pairs based on a Euclidean distance. The RNN encoder may be implemented as part of a Turbo Autoencoder (TurboAE) encoder.

First claim

What is claimed is:

1. A symmetric interleaver for a Convolutional Neural Network (CNN) or Recurrent Neural Network (RNN) encoder, comprising:
an input interface configured to receive an input block of K elements;
an output interface configured to output an output block of K elements; and
a controller configured to:
identify a prime relationship in which a parameter δ and K have 1 as the only common factor;
interleave, based on the prime relationship, elements of the input…

Assignees

• Samsung Electronics Co Ltd

Inventors

• Yildiz Hikmet
• Hatami Homayoon
• Bae Jung Hyun

Classifications

• G06N3/0464

  Physics · mapped topic

• G06N3/09

  Physics · mapped topic

• G06N3/0442

  Physics · mapped topic

• G06N3/0455 Primary

  Physics · mapped topic

• H03M13/2732

  Electricity · mapped topic