Researchers from Beijing University of Posts and Telecommunications (BUPT) and Pengcheng Laboratory introduced the Motion-State-Series Trajectory Prediction and Online Anomaly Detection (MSSTP-OAD) system to combat GPS spoofing.
By training a stacked LSTM on flight logs to forecast short-term motion, the system detects anomalies in real time through a two-stage ensemble decision process. Simulation tests on 30,000 flight segments achieved trajectory prediction accuracy above R = 0.99 and spoofing detection with 98.4 percent accuracy, while reducing off-course distance by 26 percent compared with baseline methods.
Field trials with radio spoofers are planned, and future work includes fusing barometer and magnetometer data, quantization-aware training for lightweight firmware, and distributed fleet-level consensus voting.
Meanwhile, Dr. Fangtao Yang and colleagues surveyed the fast-growing field of covert UAV communications. Because line-of-sight wireless links are highly exposed, covert approaches aim to conceal transmissions entirely, preventing interception attempts.
The review covers uncertainty-based methods such as background noise, power, and channel variation, as well as advanced wireless techniques using multiple antennas, reconfigurable intelligent surfaces, mmWave, Terahertz, and AI.
Applications span 6G-enabled UAV networks, IoT systems, visible light communications, and ultra-reliable low-latency links. Key future challenges include energy efficiency, cellular integration, spectrum prediction, and heterogeneous interference.
A third study, led by Professor Tao Song at Beijing Institute of Technology, addresses the complex task of UAV landings on moving ships. The team developed a flight-envelope-constrained control strategy that predicts ship roll and pitch in real time using a Sliding Data Window Auto-Regressive model. This enables identification of "quiescent periods" when landings are safest.
A barrier-function-based non-singular terminal sliding mode controller, integrated with a fixed-time disturbance observer, ensures rapid convergence, suppresses overshoot, and enhances robustness against disturbances. The approach balances precision and safety in touchdown maneuvers, with future work targeting vision-based landing in GNSS-denied conditions.
Together these advances outline a multi-pronged strategy for UAV resilience: anomaly detection to expose spoofing, covert communications to shield wireless links, and predictive control to enable safe landings in dynamic maritime environments. Such innovations will be crucial as UAVs assume wider roles in civilian and defense applications.
Research Report:Prediction-based trajectory anomaly detection in UAV system with GPS spoofing attack
Research Report:A survey of covert UAV communications
Research Report:Flight envelope constrained UAV shipboard landing control within an identified quiescent period
Related Links
Beijing University of Posts and Telecommunications
UAV News - Suppliers and Technology
| Subscribe Free To Our Daily Newsletters |
| Subscribe Free To Our Daily Newsletters |
