Chun Yang and Andrey Soloviev, QuNav

View Abstract Sign in for premium content

Abstract:

With a single antenna, it is rather difficult to discern an authentic GPS signal from a spoofing signal merely based on the time-domain waveforms especially when the two signals are off only slightly in time, frequency, and phase. The root-cause of vulnerability of all conventional GPS receivers to power-dominant creep-in spoofing attacks is their “myopic” architecture, which has only three correlators and cannot “see” far beyond ±1.5 chips until too late. When the spoofing signal is on top of the authentic signal, their summation moves the code and carrier error discriminators in a way similar to multipath except that one is location-dependent (multipath) while the other is persistent (spoofing). The fact is that the presence of a spoofing signal cannot eradicate the existence of an authentic signal even though it may overpower the latter. If a spoofing signal always aligns itself to the authentic signal, no harm is done. When a spoofing signal captures a receiver and starts to pull away, two correlation peaks will emerge in the time-frequency domain, one belonging to the spoofing signal and the other to the authentic signal. This fact has been used in the past for spoofing detection. In this paper, we go one step further to develop an all signal acquisition processing (ASAP) scheme that not only mitigates the effect of spoofing but also estimates the spoofing signal for spoofing intent analysis. This paper first details a spoofing simulator implementation that enables high-fidelity software-based injection of spoofing signals into live-sky GPS RF signal data. The ASAP scheme is then described, which provides a range protection level better than 1 chip (300 m) with the conventional correlator architecture. Similarly, the signals can also be separated at the output of 20 ms correlations if off in frequency by 25 Hz. Higher range protection level and tighter spectral separation are possible for a modified architecture. Simulation results are presented to illustrate the functionality and performance of spoofing detection, estimation, and mitigation.