yuting-li/4th_Year_Project_FINAL

Exploration of Deep Reinforcement Learning algorithms and their applications in network packet routing

Network Packet Routing with Deep Reinforcement Learning

IMPORTANT NOTE: Programs were only tested on Linux, compatibility with Windows is not expected.

Python 3.7 and above is recommended.

It is recommended to install all libraries and run the Python program within an env.

This can be done by running python3 -m venv <env_name>

To enable this environment run source <env_directory>/bin/activate

Libraries required for Python 3 RL implementations:

Numpy Torch Tensorflow gym

Libraries for the Python Map program:

gmplot3 bs4

In order to run our reinforcement learning implementations, run

python RunRL.py -h It takes command line arguments to select the network, but defaults to germany50 so if python RunPolicyGradients.py is ran it will by default run germany50 and not render

to run another network you would run with the -n flag for example: python RunRL.py -n newyork to render it you would run with the -r flag: python RunRL.py -n atlanta -r