A Robust GNSS LOS/NLOS Signal Classifier

Research output: Contribution to journalArticlepeer-review

91 Scopus citations

Abstract

GNSS signal classification to LOS and NLOS signals is of great value for conventional ranging-based and shadow matching algorithms. The most common attribute for performing this classification is the signal strength. Alas, such classification is often insufficient, in particular, in urban environments. In this paper, we present a novel approach for LOS/NLOS classification utilizing supervised machine learning algorithms. Provided with a sufficiently large labeled training set, the proposed approach is able to predict with high certainty (>85 percent) the satellites' visibility status in dense urban regions. This achievement was possible due to the vast raw measurements supplied for the algorithm and using sophisticated feature-selection techniques.

Original languageEnglish
Pages (from-to)429-442
Number of pages14
JournalNavigation, Journal of the Institute of Navigation
Volume63
Issue number4
DOIs
StatePublished - 1 Dec 2016

Fingerprint

Dive into the research topics of 'A Robust GNSS LOS/NLOS Signal Classifier'. Together they form a unique fingerprint.

Cite this