Uncoupled Accelerometer Based GNSS Spoof Detection for Automobiles Using Statistic and Wavelet Based Tests

Abstract

Low cost IMUs have shown potential for use in spoof detection of dynamic vehicles. IMUs represent independent sources of measurement that can be corroborated with accelerations derived from GNSS measurements. This work looks at statistical methods used in previous work in aircraft and investigates their use in spoof detection for cars. Although signatures in dynamics for cars are not as strong as they are in aircraft, there is agreement between both sources of measurement. The statistical methods previously used, however, are not yet robust to factors in the automobile environment contributing to false alarms. A method focused on the frequency content using wavelet coherence is introduced that has the potential to greatly simplify the operations necessary to compare acceleration signatures.