RAIM Algorithms Analysis for a Combined GPS/GALILEO Constellation

Abstract

How far a user can rely on his navigation system is
\na central question for safety of life applications like air
\nnavigation especially in approach phases. For Oceanic,
\nEn route or Non Precision Approach phases, the integrity
\nrequirements as defined by the ICAO (International Civil
\nAviation Organization) should be fulfilled by the future
\nGalileo Safety of Life Service.
\nThis paper presents the performances of RAIM algorithms
\nusing a covariance matrix of a single frequency absolute
\npositioning receiver noise calculated using one year
\nmeasurement data.
\nIn the configuration of combined GPS/GALILEO constellation,
\nthe user will have the possibility to track at
\nleast 10 satellites at the same time. This high availability
\nof satellites will provide a high availability of RAIM
\nalgorithms.
\nThe original approach used in this paper is to use the IPRE
\n(Instantaneous Pseudo Range Error) developed in [1] as
\nthe input parameter of the RAIM algorithms. This concept
\nprovides a generalized covariance matrix of pseudo range
\nnoise taking into account correlations of pseudo range
\nerrors with close elevation and azimuth angles. Thanks
\nto a Cholesky decomposition, it is always possible to use
\nthe classical χ2 distribution to obtain the fault detection
\nthreshold. The advantage of generalizing the RAIM
\nmethods is not only in the simplicity of the algorithm, but
\nit is also in its efficiency thanks to lower protection levels
\nobtained.