Design and Analysis of a Dual Antenna Vector Tracking Software Defined Receiver for Robust Navigation and Mitigation of Spoofing Threats

Abstract

Global navigation satellite systems (GNSS) remain the best way to maintain a global position, velocity, and timing (PVT) solution for autonomous systems. However, GNSS is susceptible to interference and degradation. One of the most nefarious forms of interference is spoofing attacks, in which a bad actor broadcasts a false GNSS signal to deceive a GNSS user into trusting false navigation states. This can be a simple rebroadcast of revived signals, or it can be a complicated directed attack against a single target. Although the received signal parameters appear legitimate, the angle of arrival of the incoming signals can identify whether the signal is authentic or spoofed. Typically a planar antenna array can identify the angle of arrival of a signal in three dimensional space to determine if it is authentic or not. A linear array can look for phase consistencies to classify spoofing signals but cannot determine a full look direction vector. This work introduces a dual antenna vector tracking receiver, which is a software defined radio that is deeply coupled with a small antenna array, and uses navigation state feedback to synthesize larger antenna arrays to achieve results of a larger planar array with a fraction of the hardware requirements. The synthetic array uses estimates of receiver and emitter motion to get correlations at future points in time at set separations so that the correlators are phase consistent with an instantaneous capture of a larger array. The correlation observations are used to determine angle of arrival and classify spoofing signals. Additionally, the classification algorithm is adapted such that it functions with no initial knowledge of the receiver PVT, so that it will work if it comes online in a spoofing environment.

The Dual Antenna Vector Tracking receiver (DAVT) is tested for its direction finding capabilities using commercial simulations and live sky tests, and compared against the performance of a small planar array. It was found that the DAVT was able to accurately provide angles of arrival when the receiver was moving at speeds greater than 5 m/s and had a C/N0 greater than 30 dB-Hz. Additionally, it was found that the DAVT could perform at levels near that of a four element planar array on live sky tests and critically provided similar test statistics that would be used in the spoofing signal classification scheme.

The spoofing classification scheme was tested using a commercial simulation tool to create a spoofing scenario with multiple cooperative emitters. The DAVT was able to correctly identify the spoofing signals and the authentic signals using Angle of Arrival (AOA) with the only initial knowledge being the ephemerides. After the classification, a mitigation scheme was introduced and tested on simulated data that exploits the deterministic nature of spoofing signals. A digital beamformer is used to isolate spoofing signals who are actively interfering with authentic signals. The signal parameters of the spoofing signal are estimated and successive interference cancellation is used to subtract the estimated signal from the data stream. The spoofing free data streams are able to use the spatial degrees of freedom to instead amplify the authentic signal. This mitigation algorithm showed that beam forming improved positioning performance in cases where the spoofing signals had up to a 20 dB power advantage over the authentic signals.