/Antenna-Selection-and-Beamforming-with-BandB-and-ML

Machine learning accelerated Branch and Bound for Joint beamforming and antenna selection

Primary LanguagePython

Optimal Solutions for Joint Beamforming and Antenna Selection: From Branch and Bound to Machine Learning

Implementation of the paper "Optimal Solutions for Joint Beamforming and Antenna Selection: From Branch and Bound to Machine Learning".

Installation

Make sure that you have pip installed.

If you are on linux. Run the following command

./initial_setup.bash

Else, create the following directories

antenna_selection/data_bf/data
antenna_selection/data_rbf/data

antenna_selection/data_bf/trained_models
antenna_selection/data_rbf/trained_models

And run

pip install -r requirements.txt

Executing the Code

On the base directory run the following to make sure that the repo is added to PATH and PYTHONPATH environment variables

source activate_env.bash

Finally, you can run the proposed B&B procedure by running the following

python antenna_selection/bb_unified.py

You can train node classifier with the following

python models/dagger_multiprocess.py

Change the parameters in models/setting.py and on the individual file that you are running.