28 lines
1.5 KiB
TeX
28 lines
1.5 KiB
TeX
\section{Prior Arrangements}
|
|
|
|
The main goal for our indoor navigation system was, to reduce the initial setup time as much as possible.
|
|
|
|
As we do not rely on \docWIFI{} fingerprinting, no time consuming measurements are needed.
|
|
Instead, signal-strengths are predicted using a model, that needs the position
|
|
of each transmitter (\docAP{}, \docAPshort{}) within the building and three parameters.
|
|
For further simplification, those three parameters can be the same for all transmitters
|
|
and chosen empirically.
|
|
|
|
The system's core is based on the building's floorplan which must be known beforehand.
|
|
Using an editor, developed especially for this purpose, the provided floorplan (PDF, Images, ..)
|
|
is used to (manually) derive a new data-structure that contains each of the building's floors including
|
|
its walls, doors, stairs and transmitters (see figure \ref{fig:mapEditor}). This step consumes most
|
|
of the setup time and might be (semi-)automated in the future.
|
|
|
|
\begin{figure}
|
|
\centering%
|
|
\includegraphics[trim=99 0 0 0, clip, width=8.2cm]{editor1.png}%
|
|
\caption{One floor within a building, created by using our map-editor}
|
|
\label{fig:mapEditor}
|
|
\end{figure}
|
|
|
|
The hereby created data-structure is used to derive a graph (automatically, offline),
|
|
that is transmitted to the smartphone (once per building)
|
|
and serves as base for the transition process (see \refeq{eq:recursiveDensity}),
|
|
modelling pedestrian walking behaviour and described within the next section.
|