Hypothesis testing methods to detect spoofing attacks: a test against the TEXBAT datasets

Abstract

The hazardous effects of spoofing attacks on the global navigation satellite system (GNSS) receiver are well known. Technologies and algorithms to increase the awareness of GNSS receivers against such attacks become more important and necessary. We present the validation of two statistical spoofing detection methods, namely the Chisquare goodness of fit (GoF) test and the Sign test applied to pairwise correlator differences, for each satellite tracked by the receiver. The test bench for the algorithms, both implemented in a software receiver, is the public database produced by the University of Texas at Austin, which reproduces various representative cases of spoofing attacks (the so-called TEXBAT). The algorithms show a very promising capability of detecting the attack, in particular when an aggregate decision is taken based on a joint detection upon all the tracked satellites. Furthermore, the GoF test appears also reliable in dynamic conditions and in case of a huge power advantage of the spoofing signal. The response of the receiver to the attacks confirms the spoofing signal represents an ''extraneous agent'' which, before taking control of the receiver, can be recognized by properly combined strategies of signal quality monitoring.

Situational Awareness: Mapping Interference Sources in Real-Time Using a Smartphone App

“…As better detailed in Section 4, we propose a new simple PSD-based metric, namely the total energy of error, to be applied to the pre-correlation samples and a Signal Quality Monitoring (SQM) algorithm, namely the Chi-square Goodness of Fit (GoF) statistical test [43], to the post-correlation samples. The effectiveness of the GoF test to detect continuous wave (CW) interferers and spoofing attacks was proved in [44,45,46], while its validity with jamming has not been fully investigated yet. In this paper the selected algorithms, included the GoF test, has been analyzed to deal with jamming signals and properly modified to target a real-time Android based implementation, which is the main focus of this work.…”

“…On the contrary, in a critical scenario where interference distorts the correlation, the p -value assumes smaller values. A threshold mechanism is used to decide on the binary hypothesis, set on the basis of the ‘significance level’ of the test [44]. For a thorough theoretical description of the GoF statistical test, the reader can refer to [43,44].…”

“…A threshold mechanism is used to decide on the binary hypothesis, set on the basis of the ‘significance level’ of the test [44]. For a thorough theoretical description of the GoF statistical test, the reader can refer to [43,44]. Hereafter, the in-laboratory calibration phase as a function of the C/N 0 signal ratio in nominal conditions and the on-field test results are presented in Section 5.2 and Section 6.3 respectively.…”

“…The third in-laboratory test aimed to calibrate the reference distribution function of the Chi-square GoF Test. In a previous work [44,45], the reference distribution function was computed and stored during the calibration phase executed in a portion of “clean” signal; in this way, the detection method was calibrated every time a channel starts tracking. The approach [44,45] shows a drawback in the real-time application because it implies repeating the calibration for each channel and for each time the detector starts monitoring, using a portion of non-interfered signal.…”

“…In a previous work [44,45], the reference distribution function was computed and stored during the calibration phase executed in a portion of “clean” signal; in this way, the detection method was calibrated every time a channel starts tracking. The approach [44,45] shows a drawback in the real-time application because it implies repeating the calibration for each channel and for each time the detector starts monitoring, using a portion of non-interfered signal. Therefore, in NGeneApp the calibration phase was performed in the laboratory and the reference distributions are loaded from static memory every time the receiver is switched on.…”