This repositiry is about using deep neural networks to build MIMO(multiple-input and multiple-output) detectors.
The architecture of DetNet is referenced by https://arxiv.org/abs/1706.01151.
Step1: adjust the hyperparameters in parameters.py
as needed.
Step2: use generate_data.py
to synthesize the validation dataset.
Step3: let's start training.
Main package | Version |
---|---|
Python | 3.7.4 |
Matplotlib | 3.1.1 |
Numpy | 1.16.5 |
TensorFlow | 1.14* |
*If you wanna run on TensorFlow 2, add tf.compat.v1.disable_eager_execution()
in main.py
.