Evaluating GPS Data in Indoor Environments

Abstract

With the latest generation of ultra-sensitive GPS-receivers, satellite signals can often be picked up even indoors, resulting in (inaccurate) indoor GPS-localization. A covered position will therefore no longer be characterized by the absence of satellite signals, creating the need for another way of categorizing this data as potentially inaccurate. This paper describes the use of GPS-based localization in an indoor environment. Only high level, generally available, GPS-data (NMEA-0183 GNSS-subset) are taken into account. Applications of ubiquitous location awareness, where the use of several RTLS (Real Time Location System) combinations is feasible, may benefit from this information to discriminate between GPS and other available localization data. A quality indicating parameter is readily available in GPS-data; the DOP (Dilution Of Precision) data field, which indicates the accuracy of the GPS localization based on the current satellite geometry. However since in indoor environments the roof and possible overlying floors often cause more signal attenuation compared to (outer) walls or windows, the probability of a better reception of 'low' orbiting satellite signals increases, giving rise to an unjustified good horizontal DOP value. Standard NMEA-0183 GPS strings are therefore analyzed in search of other indicators for malicious GPS-data.