Study of the temporal behaviour of GPS/GALILEO NSE and RAIM for LPV200
Résumé
For many years, civil aviation has identified GNSS as an attractive means to provide navigation services for every phase of flight due to its wide coverage area. However, to do so, GNSS must meet stringent requirements in terms of accuracy, integrity, availability and continuity. To achieve this performance, augmentation systems have been developed to correct the GPS L1 C/A and to monitor the quality of the received Signal-In-Space (SIS). Different solutions exist depending on where and how the augmentation is implemented. We can distinguish ABAS (Aircraft Based Augmentation Systems), SBAS (Satellite Based Augmentation System) and GBAS (Ground Based Augmentation System). ABAS systems are very interesting as they are autonomous, meaning that no signal external to the aircraft is needed to monitor the SIS thereby reducing the dependency on a ground navaids network. It uses redundant information within GNSS constellation to provide integrity monitoring. It is important to note that unlike other augmentation systems, ABAS does not improve position accuracy. Two types of ABAS systems can be distinguished: Receiver Autonomous Integrity Monitoring (RAIM) where only GNSS information is used, Aircraft Autonomous Integrity Monitoring (AAIM) where information from other on-board sensors is also used. This paper will study the RAIM temporal behavior. Today RAIM and/or AAIM are commonly used to provide integrity monitoring for phases of flight down to Non Precision Approaches using GPS L1 C/A measurements. However, current performance associated with GPS L1 and GPS constellation is not sufficient to meet civil aviation requirements for more stringent phases of flight and in particular, the associated vertical requirements of APV and LPV200. The introduction of new signals and constellations - such as GPS L5 and GALILEO for example - will significantly increase the number of available signals and satellites, the quality of the measurements as well as the quality of constellation geometries. Thus, RAIM may provide a simple means to monitor the quality of the SIS and to reach more stringent phases of flight such as approaches with vertical guidance like APV or LPV200 operations. This possibility has been investigated by the civil aviation community during recent years and RAIM is now foreseen as an interesting candidate to provide integrity monitoring for LPV200. Different algorithms have been studied and their performance, in terms of availability, has been published. It results that RAIM alone does not appear to provide sufficient protection for LPV200 approaches. New possibilities to extend RAIM to APV and LPV200 are under study and it can be reasonably assumed that RAIM using GPS and GALILEO constellations could be used in the near future (Advanced RAIM). Therefore, it appears necessary to address the specific phenomena that will result from the combination of two different constellations in one positioning solution. The aim of this paper is to present a microscopic analysis of the temporal behavior of GPS and GALILEO Navigation System Error (NSE) and RAIM, under various constellations and signal configurations for LPV200 types of operation. Indeed, the use of different types of satellites, the constellation changes, the potential loss of frequencies may imply unexpected behavior of NSE. Thus, the impact of the introduction of this new positioning solution must be investigated so as to limit unwanted effects for aircraft flight control systems which use the GNSS computed position. To enable this study to take place, a GPS/GALILEO + RAIM receiver model has been developed. The first section is a presentation of the receiver model used for the simulations which is composed of the following different modules: an aircraft trajectory generator, the measurement error models, the receiver model, the Position Velocity Time (PVT) module and the integrity monitoring (only RAIM will be used in this study). The receiver model is based on a correlator outputs generator able to model different receiver configurations for several signals and constellations. The error models used as well as the inputs and outputs of the simulator are described. The Least Square Residuals (LSR) RAIM algorithm is also detailed. The second section describes the study case defined for our simulations including various configurations of constellations, signals and approaches as well as the spatial and temporal grid used to evaluate the behavior of NSE. The following section provides a microscopic study of the position error time behavior when using GPS and GALILEO dual frequency, and in the presence of constellation changes. Aircraft maneuvers along the approach will be considered so as to observe their impact on the error. For example, a sharp turn can result in the loss of a satellite due to the antenna pattern. The next section discusses the benefits of using special configurations with respect to the additional complexity and errors introduced. The typical example to study is the resulting benefits from using single frequency on one constellation and dual frequency on the other. On the one hand, additional satellites provide a better geometry but on the other hand single frequency measurements are far less accurate than dual frequency measurements and will therefore introduce additional errors in the positioning solution. This section also investigates the impact of unbalanced numbers of satellites in the two different constellations. Finally, the last section of this paper studies the possibility of using methods such as constellation prediction to prevent unwanted NSE behaviors.