This project applies the Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm to perform efficient routing on networks. MADDPG is a powerful reinforcement learning algorithm that provides solutions in environments where multiple agents must cooperate or compete.
The project aims to improve network routing efficiency by leveraging the robustness and adaptive nature of the MADDPG algorithm.
Before running this project, make sure you have the following software installed on your local machine:
- Python (3.6 or above)
- PyTorch
- NumPy
- Matplotlib
- OpenAI Gym
- networkx
The program can be executed by running the file MADDPG. Right now it is using the network topology provided on the file small_network.pickle.