ITS Simulation: DRL & Metaheuristic for Joint Task Handling

This project simulates joint task handling and mission processing in Intelligent Transportation Systems (ITS) using Deep Reinforcement Learning (DRL) and metaheuristic approaches.

---

1. Environment Setup

source setup.sh

2. Project Structure

.
├── src/
│   ├── DRL/                      # DRL algorithms & training scripts
│   ├── physic_definition/       # ITS environment simulation
│   ├── meta_heuristic/          # Metaheuristic methods & analysis
│   └── ...
├── configs/                     # Configuration files
├── task/                        # Output results (auto-created)
├── logs/                        # Runtime logs (auto-created)
├── run.py                       # Main entry point
├── requirements.txt             # Python dependencies
└── README.md                    # This file

---

3. Run Simulations

Run simulations with run.py:

python run.py -i <method> [--verbose] [-device <cuda_id>] [-a <analysis_mode>] [-c <comparison_mode>]

Examples:

• Run DDQN:

  python run.py -i ddqn

• Run many metaheuristics:

  python run.py -i many_metaheuristics

• Run evaluation of DDQN results:

  python run.py -i eval_ddqn

• Compare DRL and metaheuristics:

  python run.py -i meta_heuristic_proposal -c drl_and_meta_heuristic_proposal

• Run analysis mode 1:

  python run.py -i None -a 1

-a triggers statistical analysis and plotting
-device -1 uses CPU, 0 uses first CUDA GPU

---

4. Logging

All logs are automatically saved in the logs/ folder, for example:

./logs/run_log_20250806_153245.log

---

5. Plotting Style

This project uses the scienceplots package for publication-ready matplotlib styles.

---

📚 Citation

If you use this codebase in your research, please cite the following works:

📄 1. arXiv Preprint

Oranits: Mission Assignment and Task Offloading in Open RAN-based ITS using Metaheuristic and Deep Reinforcement Learning
Ngoc Hung Nguyen, Nguyen Van Thieu, Quang-Trung Luu, Anh Tuan Nguyen, Senura Wanasekara, Nguyen Cong Luong, Fatemeh Kavehmadavani, Van-Dinh Nguyen
arXiv: 2507.19712

@article{nguyen2025oranits,
  title     = {Oranits: Mission Assignment and Task Offloading in Open RAN-based ITS using Metaheuristic and Deep Reinforcement Learning},
  author    = {Nguyen, Ngoc Hung and Thieu, Nguyen Van and Luu, Quang-Trung and Nguyen, Anh Tuan and Wanasekara, Senura and Luong, Nguyen Cong and Kavehmadavani, Fatemeh and Nguyen, Van-Dinh},
  journal   = {arXiv preprint arXiv:2507.19712},
  year      = {2025},
  url       = {https://arxiv.org/abs/2507.19712}
}

---

📄 2. IEEE GLOBECOM 2025

A Metaheuristic Approach for Mission Assignment and Task Offloading in Open RAN-Enabled Intelligent Transport Systems
Ngoc Hung Nguyen, Nguyen Van Thieu, Quang-Trung Luu, Vo Phi Son, Van-Dinh Nguyen
Accepted at IEEE GLOBECOM 2025, Communication QoS, Reliability and Modeling Symposium

@inproceedings{nguyen2025metaheuristic,
  title     = {A Metaheuristic Approach for Mission Assignment and Task Offloading in Open RAN-Enabled Intelligent Transport Systems},
  author    = {Nguyen, Ngoc Hung and Nguyen, Van Thieu and Luu, Quang-Trung and Vo, Phi Son and Nguyen, Van-Dinh},
  booktitle = {Proceedings of the IEEE Global Communications Conference (GLOBECOM)},
  year      = {2025},
  organization = {IEEE}
}

---

📌 Please cite one of the arXiv preprints or the GLOBECOM paper when using this code.