098 - Metrics to Assess the Performance of GNSS Authentication Schemes

All 2nd Generation GNSSs are making use of Authentication (civil signal anti-spoofing). The potential spoofing attack vectors are known: Meaconing, record and replay, fake signal generation. Different Authentication schemes (system level) will be used and applied as countermeasures in addition to user related measures. However, a gap exists in the evaluation methodology of the robustness performance: it is unclear how the authentication methods perform against meaconing, replay and fake attacks in a quantitative way. The focus of the project is to develop a set of metrics to measure the anti-spoofing performance of GPSIII(Chimera), BDS III, QZSS. Possible authentication metrics (see J. Anderson et al. CHIMERA for GPS civilian signals, 2023) as a starting point could be System openness (keys availability), Time-to-First authenticated channel, Time-between-Authentication, bit-strength, correlation loss, authentication error-rate. For comparison, Galileo (G1G) Navigation Message Authentication and Commercial Authentication Service will be used as a reference.

The objective of the activity is to Develop, test and verify a methodology to evaluate the performance of GNSS authentication. 

The tasks to be performed shall include: 

• Technical information on authentication schemes across the 2nd generation GNSS (GPS III, BDS III, QZSS, Galileo) have to be collected and analysed in detail. In the literature, proposed metrics are considered as a starting point, mapped into mathematical descriptions. The potential of the metrics for authentication performance definition has to be evaluated. New possibilities should be proposed. The result is a set of useful evaluation algorithms for authentication robustness and performance. 
• Three spoofing cases (meaconing, replay, fake signal) for spoofing attacks are defined in a parametric way. The metrics declared useful, are in a first step tested by S/W simulations and in a second step by using a GNSS constellation simulator and a real receiver. The results for different GNSS are compared.

The main outputs of the activity will consist of: 

• Ability to define the performance and robustness of GNSS authentication in a systematic way 
• Establish a Learning curve for LEO-PNT authentication