2009 | OriginalPaper | Chapter
On Structural Identification of 2D Regression Functions for Indoor Bluetooth Localization
Authors : Rene Mayrhofer, Stephan Winkler, Helmut Hlavacs, Michael Affenzeller, Stefan Schneider
Published in: Computer Aided Systems Theory - EUROCAST 2009
Publisher: Springer Berlin Heidelberg
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
In-door localization of mobile devices is a common problem for many current and future applications, for example to control infrastructure services or for personalized in-building navigation systems. Sufficiently capable Bluetooth support is often available in off-the-shelf mobile devices such as mobile phones, which makes Bluetooth an attractive technology for cheap and widely available in-door localization systems. However, Bluetooth has been optimized to deal with effects of radio frequency transmission such as reflection and multi-path propagation. It therefore produces highly non-linear relationships between the distance of devices and their perceived signal strength. In this paper, we aim to identify these relationships for a specific dataset of 2D device positions using structural identification methods. Driven by an extended genetic algorithm, we aim to find optimal mappings in form of non-linear equations for
x
and
y
coordinates, thus producing formal regression functions.