This project is a simulation framework for comparing routing algorithms in quantum networks. It uses a sophisticated Deep Reinforcement Learning (DRL) agent to find optimal paths for entanglement distribution and benchmarks its performance against classical algorithms like Dijkstra's and Random Walk across various network topologies.

• 🚀 Getting Started
  • Prerequisites
  • Installation & Execution
• 📈 Viewing Results
• 🔬 Core Technology
• 🔧 Running a Single Experiment
• 📁 Project Structure
• 📄 License

To get the project running, follow these steps.

• Python 3.9+
• Git

1. Clone the repository

   git clone <repository_url>
   cd <repository_folder>

2. Install dependencies

   pip install -r requirements.txt

3. Navigate into the source directory

   cd src

4. Make the experiment script executable (only needed once)

   chmod +x run_experiments.sh

5. Run the full experiment suite

   ./run_experiments.sh

After the script finishes, all logs, comparative plots, and summary reports will be available in the outputs/ directory at the root of the project.

Note: The run_experiments.sh script automatically calls the analysis script upon completion.

This project integrates several modern technologies to achieve its goals:

• DRL Agent:

  • Noisy Dueling DQN: For more stable and efficient Q-value estimation and learned exploration.
  • Prioritized Experience Replay: To focus training on the most informative transitions.

• Simulation:

  • Python: Core language for simulation logic.
  • NetworkX: For creating, manipulating, and studying complex network structures.

• Machine Learning:

  • PyTorch: Framework used to build and train the DRL agent.

• Data Analysis:

  • pandas: For manipulation and aggregation of experiment logs.
  • Matplotlib & Seaborn: For generating high-quality comparative plots and visualizations.

To run a specific simulation without executing the full suite, use main.py from within the src directory:

# Ensure you are in the 'src' directory
cd src

# Example: Run DRL on the 'nsfnet' topology for 2000 episodes
python main.py --topology nsfnet --algorithm drl --episodes 2000

The project is organized into a clean src and outputs structure to separate source code from generated files.

.
├── src/
│   ├── main.py                 # Main script to run a single experiment
│   ├── run_experiments.sh      # Shell script to automate all experiments
│   └── ...                     # All other source code files
└── outputs/
    ├── logs/                   # Directory for experiment CSV logs
    └── analysis_results/       # Directory for plots and reports

This project is licensed under the terms of the license agreement included in the LICENSE file.