Authors : Rahul Singh; Satish Kumbhar

Volume/Issue : Volume 10 - 2025, Issue 5 - May

---

Google Scholar : https://tinyurl.com/y5wn6uph

DOI : https://doi.org/10.38124/ijisrt/25may896

Google Scholar

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Note : Google Scholar may take 15 to 20 days to display the article.

---

Abstract : This study presents an exploration into the use of Reinforcement Learning (RL), specifically Deep Q-Networks (DQN), for autonomous drone navigation within complex, obstacle-rich environments. Utilizing Microsoft’s AirSim simulator and an open-source DRL integration framework (AirsimDRL), the research trains a drone to intelligently reach target destinations while avoiding collisions. The agent interacts with a dynamic simulated world, learning optimal control strategies from scratch. The study aims to bridge the gap between traditional UAV path planning and intelligent, learning- based navigation systems, laying the foundation for real-world autonomous drone applications.

Keywords : Reinforcement Learning (RL), Deep Q-Network (DQN), Autonomous Drone Navigation, AirSim Simulator, UAV, Deep Reinforcement Learning (DRL), Obstacle Avoidance, Smart Mobility, AI-based Navigation, Flight Path Optimization.

References :

1. Sunghoon Hong. AirsimDRL. GitHub Repository: https://github.com/sunghoonhong/AirsimDRL
2. Microsoft AirSim Documentation. https://microsoft.github.io/AirSim
3. Mnih, V., et al. (2015). Human-level control through deep reinforcement learning. Nature, 518, 529–533.
4. Lillicrap, T., et al. (2015). Continuous control with deep reinforcement learning. arXiv:1509.02971
5. OpenAI Spinning Up. An Educational Resource on RL. https://spinningup.openai.com
6. Mnih, V., Kavukcuoglu, K., Silver, D., et al. (2015). Human-level control through deep reinforcement learning. Nature, 518(7540), 529–533. https://doi.org/10.1038/nature14236
7. Lillicrap, T.P., Hunt, J.J., Pritzel, A., et al. (2016). Continuous control with deep reinforcement learning. arXiv preprint, arXiv:1509.02971. https://arxiv.org/abs/1509.02971
8. Shah, S., Dey, D., Lovett, C., Kapoor, A. (2017). AirSim: High-Fidelity Visual and Physical Simulation for Autonomous Vehicles. Field and Service Robotics (FSR). https://github.com/microsoft/AirSim
9. Zhang, W., Liu, H., Chen, Y., et al. (2023). RL-based obstacle avoidance for UAVs in urban environments.
   Journal of Aerial Robotics, 12(1), 34–45. Sutton, R.S., & Barto, A.G. (2018). Reinforcement Learning: An Introduction (2nd Edition). MIT Press. http://incompleteideas.net/book/the-book-2nd.html
10. Zhu, H., Liu, J., & Zhao, Y. (2022). Reinforcement Learning with Model Predictive Control for Autonomous Navigation. In Proc. of IEEE Conference on Decision and Control (CDC). Chen, M., Zhao, L., & Zhou, R. (2024). Using transfer learning to improve data efficiency in RL for UAVs. Journal of Intelligent Robotic Systems, 110(4), 567–580.
11. OpenAI. (2023). Spinning Up in Deep RL. A beginner-friendly introduction to deep reinforcement learning.
    https://spinningup.openai.com
12. Hong, S. (2022). AirsimDRL: Deep Reinforcement Learning in AirSim.
    GitHub Repository. https://github.com/sunghoonhong/AirsimDRL
13. Kumar, K.R.P., Bhaskar, S.N.R., & Rao, A.B.V. (2023). A comprehensive review of reinforcement learning in robotics: Trends and future directions. Journal of Robotics and Autonomous Systems, 123, 50–62.
14. Abbeel, P., & Levine, S. (2016). Deep reinforcement learning in robotics. Annual Review of Control, Robotics, and Autonomous Systems.
15. Mahmood, A. R., Korenkevych, D., Komer, B., & Bergstra, J. (2018). Benchmarking reinforcement learning algorithms on real-world robots. Conference on Robot Learning (CoRL).
16. Hester, T., Vecerik, M., Pietquin, O., et al. (2018). Deep Q-learning from Demonstrations. AAAI Conference on Artificial Intelligence.
17. Sutton, R.S., & Barto, A.G. (2018). Reinforcement Learning: An Introduction. MIT Press.