Performance Evaluation of Channel Decoding With Deep Neural Networks
October 30, 2017 ยท View on GitHub
Paper about the deep neural network decoder.
Introduction
We propose three types of NND, which build upon multi-layer perceptron (MLP), convolution neural network (CNN) and recurrent neural network (RNN).
Steps of Code
- Generate the training samples saved at
data/underRNN/.
cd RNN/
python get_data.py
- Run the deep neural network, BER is saved at
result/.
python mlp.py
python cnn.py
python lstm.py
- Plot the figure.
python plot.py