Session A2: GNSS Security: Interference, Jamming, and Spoofing 1

Towards a Trustworthy Position Solution; OSNMA Authentication and the use of Multi-Constellations
Annemarie van Zwol and Heiko Engwerda, Royal Netherlands Aerospace Centre
Location: Beacon A

Autonomously operating vehicles like cars and Unmanned Aerial Vehicles rely on the availability of a precise and trustworthy position velocity time (PVT) solution. When using Global Navigation Satellite Systems (GNSS), the user is depending on external signal sources and messages, which can be subject to distortions. In order to provide an additional protection layer, the Galileo Open Service Navigation Message Authentication (OSNMA) has been developed. OSNMA can help to authenticate the Galileo INAV messages. Different strictness levels of OSNMA implementation into the PVT engine can be considered, and the strictness of the implementation can affect the number of measurements available to calculate the PVT.
The use of authenticated signals can improve the security of the input to the PVT engine, but this does not guarantee the integrity of the PVT solution. To this end, algorithms such as Advanced Receiver Autonomous Integrity Monitoring (ARAIM) can be leveraged. The availability of GNSS and the corresponding protection levels depend on the local user environment, with signals being obscured by buildings or trees. By comparison of the Dilution of Precision (DOP) and protection levels obtained for the different authentication restraints, the effect of OSNMA on the integrity bounds of the positioning solution can be evaluated. Previous tests have shown that the use of a strict OSNMA implementation, where only authenticated Galileo satellite signals are used to calculate a PVT, can limit the availability of ARAIM solutions. The decrease in availability of ARAIM solutions was especially observed for measurements in urban canyon environments. For comparison the same dataset was processed with a loose implementation where only signals for which OSNMA authentication failed were excluded. Since only signals for which authentication failed were excluded, the remaining dataset can still contain satellite signals from constellations other than Galileo. This positively affected the number of observations available as input to the PVT engine, however part of the security created by performing signal authentication was lost by introducing satellite signals from constellations that cannot be authenticated.
In this paper a consistency check is proposed as a safety measure that can offer a first level of protection when using a mix of OSNMA authenticated Galileo signals and signals from other constellations. The consistency check calls for the calculation of the PVT solution from a subset of measurements containing only the OSNMA authenticated signals, calculation of the PVT solution from a subset of measurements containing only signals from the unauthenticated constellation and calculation of the PVT solution from the two datasets combined.
The authors have developed an OSNMA implementation in Matlab. This implementation allows for the study of the different steps of the authentication process and can provide insight to intermediate outputs and flags. This tool is combined with an adaptation of the MAAST implementation of ARAIM such that the combined processing chain can be studied. In addition the proposed consistency check is added.
A number of combined Galileo and GPS datasets is collected in different user environments. The proposed processing chain, including the consistency check, is performed on the collected datasets. To determine the effectiveness of the proposed consistency check GPS spoofing attacks are simulated and added to the recorded dataset. The new dataset is then processed again and results are compared.