Purdue · West Lafayette, IN

Purdue University

Game-theoretic and learning-based foundations for secure, fair, and decentralized coordination among airspace users.

Stackelberg gamesMulti-agent learningCollusion detectionFair deconfliction

Purdue’s team, led by Vijay Gupta with co-investigator Mahsa Ghasemi, develops learning and game-theoretic foundations for securing multi-UAV operations. At the individual-aircraft level, the team works on rapid detection of faults and anomalies in vehicle dynamics. At the system level, it develops learning algorithms for security games among multiple UAV owners, methods to detect collusion, digital-twin-based operation, robust information gathering, active communication in multi-agent reinforcement learning, and fair strategic deconfliction. Purdue collaborates with GWU.

Research thrusts

  1. 01

    Decentralized leader-follower learning

    Learning algorithms for Stackelberg games among UAV operators that converge without a trusted, information-rich central coordinator, preserving the privacy of each operator’s costs.

    UC-07

  2. 02

    Rational and robust learning

    Quantifying how much an agent can gain by deviating from a prescribed learning algorithm, and designing algorithms that remove that incentive.

  3. 03

    Collusion detection

    A statistical framework that detects covert coordination among agents from observed actions alone, with sample-complexity guarantees.

    UC-07

  4. 04

    Trustworthy abstractions and digital twins

    A certifiable, convex test of how faithfully a simplified model bounds the behavior of a real nonlinear system under unknown disturbances.

  5. 05

    Information gathering for intrusion detection

    Game-theoretic sensor placement, scheduling, and target coverage with provable guarantees, and robust selection of information sources.

  6. 06

    Active communication and fair deconfliction

    Multi-agent learning that decides what information aircraft should share, and fair multi-user dueling bandits for strategic deconfliction among operators.

    UC-02with GWU

Figures

From the research

Training curves showing rewards converging for two companies and collisions falling while successes rise.
GWU · PurdueTwo delivery companies, each training its own attention-based policy, converge to co-existing deconfliction strategies in shared airspace.

Demos

See Purdue's work in action

All demos

GWU · PurdueUC-02Simulation

Two delivery fleets sharing the sky

BlueSky simulation of drones from two delivery companies over Frisco, Texas, each fleet flying its own attention-based multi-agent reinforcement-learning deconfliction policy.

Use cases

Where Purdue contributes

People

Researchers and students

Postdoctoral researchers2

  • Shivam Bajaj
  • Ashwin Verma

PhD students3

  • Maheed Hatem Most Ahmed
  • Pranoy Das
  • Hyeong-Tae Kim

Publications

Publications

12 published or accepted papers

  1. A Survey of Security Challenges and Solutions for UAS Traffic Management (UTM) and Small Unmanned Aerial Systems (sUAS)

    I. Sharifi, M. Ghazanfari, A. Taye, P. Wei, M. Ahmed, H. T. Kim, M. Ghasemi, V. Gupta, B. C. Ward, X. Koutsoukos, G. Biswas, et al.

    AIAA SciTech Forum, 2026

    ConferenceGWUPurdueVanderbilt

  2. Distributed Equilibrium-Seeking in Target Coverage Games via Self-Configurable Networks under Limited Communication

    J. Bhargav, Z. Xu, V. Tzoumas, M. Ghasemi, S. Sundaram

    IEEE Conference on Decision and Control (CDC), 2026Accepted

    ConferencePurdue

  3. Finite Time Convergence Analysis of Policy Gradient Learning in Stochastic Stackelberg Games

    P. Das, M. A. Zaman, V. Gupta

    IEEE Transactions on Automatic Control, 2026Accepted

    JournalPurdueUC-07

  4. Iterative Stackelberg Equilibrium Finding for Linear Quadratic Differential Games

    B. Nortmann, P. Das, V. Gupta, T. Mylvaganam

    Automatica, vol. 183, 112648, 2026

    JournalPurdue

  5. Learning Structured Set Utility Functions with Contrastive Element Representations

    J. Bhargav, Y. Chang, M. Ghasemi, S. Sundaram

    Transactions on Machine Learning Research (TMLR), 2026

    JournalPurdue

  6. Multi-User Dueling Bandits: A Fair Approach using Nash Social Welfare

    M. H. Ahmed, M. Ghasemi

    Transactions on Machine Learning Research (TMLR), 2026

    JournalPurdueUC-07

  7. Robust Information Selection for Hypothesis Testing with Misclassification Penalties

    J. Bhargav, S. Sundaram, M. Ghasemi

    IEEE Transactions on Automatic Control, 2026

    JournalPurdue

  8. Separation Assurance between Heterogeneous Fleets of Small Unmanned Aerial Systems via Multi-Agent Reinforcement Learning

    I. Sharifi, H. T. Kim, M. H. Ahmed, M. Ghasemi, P. Wei

    IEEE International Conference on Intelligent Transportation Systems (ITSC), 2026

    ConferenceGWUPurdueUC-02

  9. Online Laplacian-Based Representation Learning in Reinforcement Learning

    M. H. Ahmed, J. Bhargav, M. Ghasemi

    International Conference on Machine Learning (ICML), 2025

    ConferencePurdue

  10. Reinforcement Learning from Multi-level and Episodic Human Feedback

    M. Q. Elahi, S. Oguchienti, M. H. Ahmed, M. Ghasemi

    Learning for Dynamics & Control Conference (L4DC), 2025

    ConferencePurdue

  11. Sensor Scheduling in Intrusion Detection Games with Uncertain Payoffs

    J. Bhargav, S. Sundaram, M. Ghasemi

    Learning for Dynamics & Control Conference (L4DC), 2025

    ConferencePurdue

  12. Rationality of Learning Algorithms in Repeated Normal-Form Games

    S. Bajaj, P. Das, Y. Vorobeychik, V. Gupta

    IEEE Control Systems Letters, 2024

    JournalPurdue