Multi-Constellation DGNSS - A Time to Revise Definitions

Abstract

A paper describes a novel approach to computing DGNSS solution when multi-GNSS measurement observables are available. The approach is applicable to situations when not all the GNSS observables are accompanied with DGNSS corrections. A commonly used “classical” recommendation for such cases is to discard observables which do not have corrections. However, in challenging conditions like urban canyon this recommendation might lead to a case, when such “correct” DGNSS solution might be of poorer quality than the stand-alone one. This is especially true for SBAS, as all operating systems broadcast GPS-only corrections, while user receiver might be capable of tracking signals of GLONASS, Beidou, GALILEO, NavIC in addition to GPS. The suggested in the paper approach is based on rather straightforward methodology, according to which all observables are divided by groups, depending on GNSS type and correction data availability. In the process of computing the position several new unknowns are introduced - one for every group, to estimate clock offsets associated with every group. This approach is implemented in the newest firmware of Topcon receiver boards, which is going to be released soon. The paper provides positioning algorithm's description and results, demonstrating benefits of the suggested approach over the “classical” one. In addition, a paper touches a problem of necessity to revise a definition of “DGNSS mode”, as today there is no term for designating a solution, computed with a mixture of corrected and uncorrected observables.