Mitigating Single Source Partial Spoofing by Applying Outlier Detection to a two Antenna Baseline Spoofing Detector

The spoofing detection technique, referred to as the “Baseline Separation Detection technique, compares the known, fixed antenna baseline separation against the calculated antenna baseline using GNSS measurements. On moving aircraft, the antennas are in fixed locations with respect to one another; however, the aircraft and antennas are in motion. A moving base state RTK solution between two antennas on a vehicle is combined with outlier detection to separate spoofed GNSS measurements from authentic measurements when some, but not all measurements are spoofed. When at least four satellites are available after outliers are removed, the RTK calculated antenna baseline separation is compared against the known separation. Note that authentic PRN wavefronts are unique at each antenna and the difference between these PRN wavefronts results in a spoofing-free solution that yields the true baseline distance (plus noise). A spoofer signal transmits all PRN wavefronts from the same antenna, and the same, identical wavefronts are received by both baseline antennas. With all measurements spoofed, the baseline separation solution will yield zero (plus noise). A partially spoofed baseline solution will not be self-consistent since the measurements are a combination of authentic and spoofing wavefronts. Authentic measurements fit the true baseline solution while spoofed measurements fit a zero baseline. An outlier detection strategy allows separation of measurements into two self-consistent groups if enough measurements are available and in turn allowing a baseline comparison to be done on a group with at least four measurements. This paper demonstrates this method using simulated data from a Spirent GSS7000 simulator. The data was processed with the off-the-shelf RTKLIB software. The results show that outlier detection is a powerful aid to baseline spoofing detection when a solution is partially spoofed.