# Deep-Reinforcement-Learning-for-Dynamic-Spectrum-Access

**Repository Path**: helloMRDJ/Deep-Reinforcement-Learning-for-Dynamic-Spectrum-Access

## Basic Information

- **Project Name**: Deep-Reinforcement-Learning-for-Dynamic-Spectrum-Access
- **Description**: Using multi-agent Deep Q Learning with LSTM cells (DRQN) to train multiple users in cognitive radio to learn to share scarce resource (channels) equally without communication
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 1
- **Forks**: 0
- **Created**: 2020-04-02
- **Last Updated**: 2021-12-23

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Deep-Reinforcement-Learning-for-Dynamic-Spectrum-Access

## Dependencies

1. tensorflow>1.0
2. matplotlib
3. python 3.x(windows)

### To install tensorflow ,follow the link [tensorflow - windows](https://www.tensorflow.org/install/install_windows) 
I recommend Installing with Anaconda 


### To train the DQN ,run on terminal
```bash
git clone https://github.com/shkrwnd/Deep-Reinforcement-Learning-for-Dynamic-Spectrum-Access.git
cd Deep-Reinforcement-Learning-for-Dynamic-Spectrum-Access
python train.py
```

To understand the code , I have provided jupyter notebooks:
1. How to use environment.ipynb
2. How to generate states.ipynb

To run notebook,run on terminal
```bash
jupyter notebook
```
Default browser will open, just open ipynb files and run them


This work is an inspiration from the paper
```
O. Naparstek and K. Cohen, “Deep multi-user reinforcement learning for dynamic spectrum access in multichannel wireless
networks,” to appear in Proc. of the IEEE Global Communications Conference (GLOBECOM), Dec. 2017
```