Bayesian Location Identifier
draft-hoene-geopriv-bli-00

Abstract

Location Generators cannot always provide exact measures of particular locations. Instead, they estimate the location of objects. More precisely, they use filters to aggregate noisy sensor data and to calculate probability density distributions of estimated positions. In location tracking applications, typically Kalman-type, Gaussian-Sum, and Particle Filters are used. We believe that it is reasonable to use the outputs of those filters to describe a location estimate and its uncertainty, because they are the natural result of location tracking algorithms. In addition, the results of those filters can be feed into sensor fusion and decision making engines easily. Geometric representations such as polygons or ellipses might be demanded by an application. The output of filters can be converted to those application demanded shapes. However, these conversions come at the loss of precision and are not well understood scientifically. Thus, we think that transmitting filter results is a solution that is easier to implement. In this draft, we present a transmission format for PIDF-LO, which is based on the output of Kalman-type, Gaussian-Sum, and Particle Filters.

Authors

Christian Hoene
Andreas Krebs
Christoph Behle
Mark Schmidt

(Note: The e-mail addresses provided for the authors of this Internet-Draft may no longer be valid.)