Selected work

Representative Papers

Conflict-Free Flight Scheduling Using Strategic Demand Capacity Balancing for Urban Air Mobility Operations

Nov 14, 2025

In high-density urban air mobility (UAM) operations, ensuring conflict-free flight and maintaining safe separation within constrained airspace remains a critical challenge for multi-agent coordination. Method: This paper proposes a strategic, supply–demand balanced, conflict-free scheduling framework. It extends pairwise conflict avoidance to multi-agent collaborative scenarios, establishing a traffic-adaptive, robust scheduling mechanism. By integrating kinematics-driven delayed takeoff control with a distributed cooperative optimization algorithm, the framework enables dynamic temporal allocation of airspace resources. Contribution/Results: Numerical simulations and real-world UAM case studies demonstrate that the method guarantees zero collisions throughout all operations while significantly reducing total system delay. Moreover, it exhibits strong scalability under increasing traffic density. The framework thus provides a viable, safety-assured, and efficiency-oriented solution for high-density UAM operations.

0 citationsRead paper

An Experimental Study of Trojan Vulnerabilities in UAV Autonomous Landing

Oct 23, 2025

Deep learning models in urban air mobility (UAM) autonomous landing systems face emerging security threats from Trojan (backdoor) attacks, yet no systematic assessment framework exists for this safety-critical domain. Method: We propose the first backdoor threat evaluation framework tailored to UAM, built upon the DroNet architecture and validated on a custom real-world flight dataset with stealthy trigger injection. We systematically assess the vulnerability of CNN-based landing navigation models under backdoor perturbations. Results: Experimental evaluation shows that the compromised model maintains 96.4% accuracy on clean inputs but suffers a severe drop to 73.3% when triggers are activated—critically degrading attitude estimation and obstacle-avoidance decision-making. This work provides the first empirical evidence of high susceptibility of UAM perception modules to data-poisoning-based backdoor attacks, establishing both theoretical foundations and a benchmark methodology for robustness verification and defense design in aviation-grade AI systems.

0 citationsRead paper

GPS Spoofing Attack Detection in Autonomous Vehicles Using Adaptive DBSCAN

Oct 12, 2025

Autonomous vehicles are vulnerable to GPS spoofing attacks—particularly subtle, incremental deviations—that compromise localization integrity. To address this, we propose a real-time detection method based on adaptive DBSCAN. Our approach recursively updates the mean and standard deviation of displacement errors under normal operation to dynamically optimize the DBSCAN neighborhood radius ε. By fusing displacement errors from GPS, IMU, and wheel-speed sensors, we construct a density-based anomaly detection model; clean-data-driven threshold initialization further enhances early attack identification. Evaluated on the Honda driving dataset, the method achieves detection accuracies of 98.62%, 99.96%, 99.88%, and 98.38% for steering, emergency braking, overshoot, and multi-segment small-bias attacks, respectively. This significantly improves localization security and robustness against sophisticated GPS spoofing.

0 citationsRead paper

Recent publications

Latest Papers

Conflict-Free Flight Scheduling Using Strategic Demand Capacity Balancing for Urban Air Mobility Operations

Nov 14, 2025

In high-density urban air mobility (UAM) operations, ensuring conflict-free flight and maintaining safe separation within constrained airspace remains a critical challenge for multi-agent coordination. Method: This paper proposes a strategic, supply–demand balanced, conflict-free scheduling framework. It extends pairwise conflict avoidance to multi-agent collaborative scenarios, establishing a traffic-adaptive, robust scheduling mechanism. By integrating kinematics-driven delayed takeoff control with a distributed cooperative optimization algorithm, the framework enables dynamic temporal allocation of airspace resources. Contribution/Results: Numerical simulations and real-world UAM case studies demonstrate that the method guarantees zero collisions throughout all operations while significantly reducing total system delay. Moreover, it exhibits strong scalability under increasing traffic density. The framework thus provides a viable, safety-assured, and efficiency-oriented solution for high-density UAM operations.

0 citationsRead paper

An Experimental Study of Trojan Vulnerabilities in UAV Autonomous Landing

Oct 23, 2025

Deep learning models in urban air mobility (UAM) autonomous landing systems face emerging security threats from Trojan (backdoor) attacks, yet no systematic assessment framework exists for this safety-critical domain. Method: We propose the first backdoor threat evaluation framework tailored to UAM, built upon the DroNet architecture and validated on a custom real-world flight dataset with stealthy trigger injection. We systematically assess the vulnerability of CNN-based landing navigation models under backdoor perturbations. Results: Experimental evaluation shows that the compromised model maintains 96.4% accuracy on clean inputs but suffers a severe drop to 73.3% when triggers are activated—critically degrading attitude estimation and obstacle-avoidance decision-making. This work provides the first empirical evidence of high susceptibility of UAM perception modules to data-poisoning-based backdoor attacks, establishing both theoretical foundations and a benchmark methodology for robustness verification and defense design in aviation-grade AI systems.

0 citationsRead paper

GPS Spoofing Attack Detection in Autonomous Vehicles Using Adaptive DBSCAN

Oct 12, 2025

Autonomous vehicles are vulnerable to GPS spoofing attacks—particularly subtle, incremental deviations—that compromise localization integrity. To address this, we propose a real-time detection method based on adaptive DBSCAN. Our approach recursively updates the mean and standard deviation of displacement errors under normal operation to dynamically optimize the DBSCAN neighborhood radius ε. By fusing displacement errors from GPS, IMU, and wheel-speed sensors, we construct a density-based anomaly detection model; clean-data-driven threshold initialization further enhances early attack identification. Evaluated on the Honda driving dataset, the method achieves detection accuracies of 98.62%, 99.96%, 99.88%, and 98.38% for steering, emergency braking, overshoot, and multi-segment small-bias attacks, respectively. This significantly improves localization security and robustness against sophisticated GPS spoofing.

0 citationsRead paper