Return to Session D3

Session D3: GNSS Augmentation and Robustness for Autonomous Navigation

Assessing the Performance of GAST-X Architecture Using the LAMBDA Method for Ambiguity Resolution
Crislaine Menezes da Silva, Weverton da Costa Silva, Felipe Tintino Linhares de Souza, Joao Francisco Galera Monico, Daniele Barroca Marra Alves, UNESP; Glaucia Balvedi, Tim Murphy, Boeing; Susumu Saito, ENRI; Joel Wichgers, Collins Aerospace

GAST-X is an alternative architecture candidate proposal for consideration for the future Dual Frequency Multi-Constellation (DFMC) Ground Based Augmentation System (GBAS). It consists of an uplink to the airborne with raw measurements from L1/E1 and L5/E5a signals in addition to the corrections and other parameters that are already uplinked from ground stations that support the existing GAST-C/D services. The airborne user can then processes the pseudorange (PR) and carrier phase (CP) measurements to produce carrier smoothed PR measurements using multiple different types of smoothing. Divergence Free (DFree) and conventional Single Frequency (SF) smoothing are performed in parallel, the latter to support legacy users (backup modes), as GAST C/D uplink data will continue to be streamed.
For GAST-X, a primary mode for positioning uses the corrected DFree smoothed pseudoranges with a long time constant (600s), and the differential corrections are computed using the ground reference station location. The use of raw measurements provides, in addition to the precise point positioning (PPP), the possibility of using pure carrier phase position solutions such as Real Time Kinematic (RTK) Carrier Phase (CP) based solutions in future aviation applications.
The use of DFree smoothed measurements also makes the system, in general, less susceptible to ionospheric gradient threats. One of the foreseen benefits of GAST-X is the higher operational availability in equatorial areas due to the higher number of satellites in view offered by the multi-constellation scenario and the possibility to remove satellites with high gradient from the solution without compromising the integrity. Further, the backup modes can come into play during signal losses and the lower noise and multipath provided by the DFree with longer smoothing times. Nonetheless, severe scintillations conditions occurring in the equatorial areas may impact signal tracking and increase cycle slips occurrences, sometimes to many satellites at the same time, which requires the validation of GAST-X algorithm using real data from equatorial regions collected during scintillation events to demonstrate its performance.
In this paper, an assessment of performance of different DFMC GBAS architectures is carried out, using L1/E1 and L5/E5a data from a set of receivers located in Brazilian territory. The receivers belong to the INCT GNSS-NavAer network and are located in the region of Presidente Prudente (state of Sao Paulo), which is under the influence of the Ionization Equatorial Anomaly (IEA). Therefore, with high ionospheric scintillation occurrences. The Day of Year (DOY) 321/2019 was chosen having presented high ionospheric scintillations with scintillation index S4 ranging from 0.6 to 1.0.
Protection levels and positioning performance were calculated for GAST D/F and X, considering different smoothing strategies. For GAST D/F, conventional smoothing were performed using 30s and 100s, and for GAST-X, carrier smoothing with 600s and also for the entire period of analyzed data.

As GAST-X makes use of CP measurements it is necessary to estimate the carrier phase ambiguities. In the initial proposal from Murphy et al. (2022) a carrier phase estimation of the iono gradient is calculated from air and ground and a differential value is estimated. The wide-lane ambiguity is solved approximating the float ambiguity as an average of the epochs used in the smoothing process. In this work it is proposed a new approach for the ambiguity fixing: the LAMBDA method (Least square AMBiguity Decorrelation Adjustment) was used to fix the wide-lane ambiguities and compute the differential estimate of the iono gradients. The integer ambiguities were used to adjust the values of the S matrix. Both the iono gradients and the S matrix is used in calculation of the GAST-X protection levels.
Preliminary results showed that the protection levels calculated using GAST-X architecture were better than the ones calculated with GAST D/F, demonstrating the robustness of the method especially when using corrected DFree smoothed pseudoranges. However, GAST-X performance can be degraded during cycle slips occurrences, and depending on the intensity of ionospheric scintillation the backup modes can be more suitable than GAST-X.



Return to Session D3