Robust GNSS Spoof Detection using Direction of Arrival: Methods and Practice

Abstract

GNSS spoofing is a growing concern due to the increasing use of GNSS in safety and economically important applications. The foremost task to manage the GNSS spoofing threat is detection. Indeed, many potential means and measurements have been proposed for spoof detection. However, there exist no panaceas. Rather, we should rely on many different detection means to provide robust detection. One measurement useful for detection is direction of arrival (DOA), which we can get from many sources such as array and dual polarization antenna (DPA). Even with accurate DOA measurements, good monitor tests are needed to make an accurate decision or determination of spoofing. This paper develops and examines monitor tests suitable for DOA based spoof detection. Developed tests aim to have provable and conservative performance levels making them suitable for safety of life applications. Several tests are developed including signal pair comparisons based on statistical hypothesis tests, Bayesian estimates as well as comprehensive all satellites in view comparisons. On-air spoofing results from a DPA are used to validate performance. The developed techniques must detect simple cases such as when all examined signals are from a single spoofer (“all spoof” case) and challenging cases where there is a mix of genuine signals and spoofed signals (“mixed spoof” case). Another consideration is that measurements will typically be relative DOA as platform attitude may or may not be known. Solving the relative DOA problem also captures the simpler case where we have some knowledge of platform attitude (i.e. absolute DOA).