B2DQN
Code for my publication: Efficient Exploration through Bootstrapped and Bayesian Deep Q-Networks for Joint Power Control and Beamforming in mmWave Networks. Paper accepted for publication to IEEE Communications Letters.
Install / Use
/learn @bszeng/B2DQNREADME
Bootstrapped and Bayesian Deep Q-Networks
How to use
- Set the number of antennas in the base station. In
environment.pychange the lineself.M_ULAto the values of your choice. The code expects M = 4, 8, 16, 32, and 64. - Run DQN variants algorithms. Run the scripts
DQN,BoDQN,BaDQN, andB2DQN.pyin folderCodes. The result is the same as that in folderResults. - Show the results. Run the script
Results_plot.ipynbin folderResultsto showFigure 3,Figure 4, andTable IVin the paper.
