Location and navigation services based on global navigation satellite systems (GNSS) are needed for real-time high-precision positioning applications in relevant economic sectors, such as precision agriculture, transport, civil engineering or mapping. Real-time navigation users of GNSS networks have significantly increased all around the world, since the 1990s, and usage has exceeded initial expectations. Therefore, if the evolution of GNSS network users is monitored, the dynamics of market segments can be studied. The implementation of this hypothesis requires the treatment of big volumes of navigation data over several years and the continuous monitoring of customers. This paper is focused on the management of massive connection of GNSS users in an efficient way, in order to obtain analysis and statistics. Big data architecture and data analyses based on data mining algorithms have been implemented as the best way to approach the hypothesis. Results demonstrate the dynamic of users of different market segments, the increasing demand over the years and, specifically, conclusions are obtained about the trends, year-on-year correlation and business volume recovering after economic crisis periods.

Most Differential Global Positioning System (DGPS) correction formats are based on range information, and thus typical DGPS systems can be implemented only on correction message-readable or raw observable-providing devices. There is no other way to improve an already-calculated position than a ‘block shift technique’, which has a very limited applicability. This paper suggests an algorithm to project measurement correction directly to position domain data without requiring raw pseudorange data. By post-processing methodology, we evaluated the performance of our new algorithm compared to conventional DGPS, which requires raw pseudorange data; the observed difference between them was only 0·1 mm . The proposed correction projection algorithm can be used with commercial off-the-shelf receivers that provide National Marine Electronics Association (NMEA) format data. Our testing with a U-blox LEA-5H receiver resulted in a drastic reduction of horizontal Root Mean Square (RMS) error from 4·75 m to 1·09 m.