Interpreting Deepcode, a learned feedback code

TL;DR

This paper analyzes Deepcode, a deep learning-based feedback code for AWGN channels, revealing its underlying principles and providing an interpretable model that matches its performance.

Contribution

It introduces an interpretable model of Deepcode based on input influence length and non-linear dynamics, enhancing understanding of feedback utilization in learned codes.

Findings

Interpretable model achieves comparable performance to Deepcode.

Analysis reveals how feedback is employed for error correction.

Model provides insights into the non-linear dynamics of the RNN encoder.

Abstract

Deep learning methods have recently been used to construct non-linear codes for the additive white Gaussian noise (AWGN) channel with feedback. However, there is limited understanding of how these black-box-like codes with many learned parameters use feedback. This study aims to uncover the fundamental principles underlying the first deep-learned feedback code, known as Deepcode, which is based on an RNN architecture. Our interpretable model based on Deepcode is built by analyzing the influence length of inputs and approximating the non-linear dynamics of the original black-box RNN encoder. Numerical experiments demonstrate that our interpretable model -- which includes both an encoder and a decoder -- achieves comparable performance to Deepcode while offering an interpretation of how it employs feedback for error correction.