Performance Evaluation of Channel Decoding With Deep Neural Networks

TL;DR

This paper evaluates the performance of different deep neural network architectures for channel decoding in 5G, highlighting RNN's superior accuracy but higher computational cost, and identifying a saturation length limit for each network type.

Contribution

It introduces and compares MLP, CNN, and RNN decoders with the same parameters, providing insights into their performance and limitations in 5G channel decoding.

Findings

RNN achieves the best decoding performance.

RNN has the highest computational overhead.

Each neural network type has a saturation length due to learning limits.

Abstract

With the demand of high data rate and low latency in fifth generation (5G), deep neural network decoder (NND) has become a promising candidate due to its capability of one-shot decoding and parallel computing. In this paper, three types of NND, i.e., multi-layer perceptron (MLP), convolution neural network (CNN) and recurrent neural network (RNN), are proposed with the same parameter magnitude. The performance of these deep neural networks are evaluated through extensive simulation. Numerical results show that RNN has the best decoding performance, yet at the price of the highest computational overhead. Moreover, we find there exists a saturation length for each type of neural network, which is caused by their restricted learning abilities.