Evaluation and telemetry-based detection of GPS spoofing effects on UAV navigation using software-defined radio

This paper presents an experimental investigation into the vulnerabilities of autonomous drone navigation systems under GPS spoofing attacks using Software Defined Radio (SDR). A low-cost spoofing framework was implemented using HackRF One and GPS-SDR-SIM to manipulate the navigation signals of an F450 quadcopter equipped with a Pixhawk 4 flight controller and a Ublox NEO-M8N GPS module. Both static and dynamic spoofing trials were conducted in controlled outdoor environments. During normal autonomous waypoint missions, the drone demonstrated stable sensor outputs and accurate trajectory tracking. However, under spoofed GPS conditions, the drone deviated from its intended path, breached geofenced boundaries, and exhibited unstable behaviour leading to an uncontrolled descent and crash landing instead of executing the intended Return-to-Launch (RTL) response. Detailed analysis of barometric altitude, yaw angle, magnetometer, vibration data, as well as GPS indicators such as satellite count and HDOP, revealed significant anomalies during spoofed flights. To enhance interpretability, a telemetry-based detection indicator was introduced by computing deviation thresholds between GPS, IMU, and barometric readings. These statistical correlations enabled the detection of spoofing-induced anomalies with over 90% reliability, offering a low-cost, platform-independent method for UAV spoofing awareness. These deviations underscore the susceptibility of GPS-dependent UAV systems to spoofing attacks. The study further demonstrates that cross-sensor anomaly detection without requiring additional hardware can serve as an effective strategy for real-time spoofing detection. This work contributes to a lightweight, platform-independent methodology for enhancing drone resilience against GPS spoofing threats through telemetry-based sensor fusion analytics.