current state
This commit is contained in:
@@ -1,80 +1,83 @@
|
|||||||
|
|
||||||
\section{Experiments}
|
\section{Experiments}
|
||||||
|
|
||||||
wir betrachten nur die fest-installierten APs die man meist anhand einer bestimmten mac-range ausmachen kann
|
\todo{
|
||||||
portable geraete von studenten, beamer, aehnliches werden ignoriert
|
alles im FHWS gebäude [korrekte groesse fuer beide gebaeude!] mit nem nexus 6
|
||||||
|
}
|
||||||
modell direkt fuer den gelaufenen pfad optimiert (also wirklich jede wifi messung direkt auf den ground-truth)
|
|
||||||
der fehler wird zwar kleiner, ist aber immernoch deutlich spürbar. das spricht dafür, dass das modell einfach nicht
|
Within all \docWIFI{} observations we only consider the \docAP{}s that are permanently installed
|
||||||
gut geeignet ist.
|
within the building. Temporal and movable transmitters are ignored as they might cause estimation errors.
|
||||||
|
|
||||||
outdoor fehler kann gemnindert werden mit z.B.
|
|
||||||
nicht nur die APs nehmen die ich sehe, sondern auch die, die ich sehen müsste
|
|
||||||
dann wird klar, dass es nicht gut passt. allerdings ist das auch gefaehrlich
|
|
||||||
[nicht immer tauchen alle APs im scan auf] und welchen fehler bzw. welche
|
|
||||||
dBm zahl nimmt man fuer fehlende APs an? das ist eine hardware-frage.
|
|
||||||
|
|
||||||
ein workaround fuer steigende fehler durch optimierung koennte sein,
|
|
||||||
dass man nicht jeden AP einzeln optimiert, sondern das gesamtsystem.
|
|
||||||
und auch nicht den dB fehler, sondern 'die wahrscheinlichkeit an dieser stelle zu sein'
|
|
||||||
bzw: 'die wahrscheinlichkeit aller anderen positionen minimieren' dass keine
|
|
||||||
fehl-positionierungen [wie outdoor] mehr stattfinden. allerdings ist das
|
|
||||||
problematisch da man auch hier entscheiden müsste wann ein AP nicht mehr
|
|
||||||
sichtbar ist etc.
|
|
||||||
|
|
||||||
walk1 hat eine issue kurz bevor man zur tuer zum hoersaalgebaude reingeht
|
|
||||||
je nach resampling killt dieser wlan error evtl alle partikel!
|
|
||||||
|
|
||||||
|
|
||||||
optimierungs input: alle 4 walks samt ground-truth
|
|
||||||
dann kommt fuer die 4 typen [fixed, all same par, each par, each par pos]
|
|
||||||
log probability 50 75, meter 50, 75
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
reines wifi eval mittels num-opt springt stark durch die gegend
|
|
||||||
d.h. das bewegungsmodell rettet uns
|
|
||||||
kann man auch testen wenn man beim particle-filter das resampling ganz aus macht
|
|
||||||
|
|
||||||
|
modell direkt fuer den gelaufenen pfad optimiert (also wirklich jede wifi messung direkt auf den ground-truth)
|
||||||
|
der fehler wird zwar kleiner, ist aber immernoch deutlich spürbar. das spricht dafür, dass das modell einfach nicht
|
||||||
|
gut geeignet ist.
|
||||||
|
|
||||||
|
optimierungs input: alle 4 walks samt ground-truth
|
||||||
|
dann kommt fuer die 4 typen [fixed, all same par, each par, each par pos]
|
||||||
|
log probability 50 75, meter 50, 75
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
% -------------------------------- optimization -------------------------------- %
|
% -------------------------------- optimization -------------------------------- %
|
||||||
|
|
||||||
% used reference measurements
|
As the signal strength prediction model is the heart of the absolute positioning component
|
||||||
\begin{figure}
|
described in \ref{sec:system} we start with the model parameter estimation (see \ref{sec:optimization}) for
|
||||||
{
|
\mTXP, \mPLE and \mWAF based on some reference measurements and compare the results
|
||||||
\centering
|
between various optimization strategies and a basic empiric choice of \mTXP = \SI{-40}{\decibel{}m} @ \SI{1}{\meter}
|
||||||
\input{gfx/all_fingerprints.tex}
|
(defined by the usual \docAPshort{} transmit power for europe), a path loss exponent $\mPLE \approx $ \SI{2.5} and
|
||||||
}
|
$\mWAF \approx$ \SI{-8}{\decibel} per floor/ceiling (made of reinforced concrete) \todo{cite für werte}.
|
||||||
\label{fig:referenceMeasurements}
|
|
||||||
\caption{
|
Figure \ref{fig:referenceMeasurements} depicts the location of the used 121 reference measurements.
|
||||||
Locations of the 121 reference measurements.
|
Each location was scanned 30 times ($\approx$ \SI{25}{\second} scan time),
|
||||||
The size of each square denotes the number of permanently installed \docAPshort{}s
|
non permanent \docAP{}s were removed, the values were grouped per physical transmitter (see \ref{sec:vap})
|
||||||
that are visible at this location,
|
and aggregated to form the average signal strength per transmitter.
|
||||||
and ranges between 2 and 22 with an average of 9.
|
|
||||||
}
|
% used reference measurements
|
||||||
\end{figure}
|
\begin{figure}
|
||||||
|
{
|
||||||
|
\centering
|
||||||
|
\input{gfx/all_fingerprints.tex}
|
||||||
|
}
|
||||||
|
\label{fig:referenceMeasurements}
|
||||||
|
\caption{
|
||||||
|
Locations of the 121 reference measurements.
|
||||||
|
The size of each square denotes the number of permanently installed \docAPshort{}s
|
||||||
|
that are visible at this location,
|
||||||
|
and ranges between 2 and 22 with an average of 9.
|
||||||
|
}
|
||||||
|
\end{figure}
|
||||||
|
|
||||||
% visible APs:
|
% visible APs:
|
||||||
% cnt(121) min(2.000000) max(22.000000) range(20.000000) med(8.000000) avg(9.322314) stdDev(4.386709)
|
% cnt(121) min(2.000000) max(22.000000) range(20.000000) med(8.000000) avg(9.322314) stdDev(4.386709)
|
||||||
|
|
||||||
\begin{figure}
|
As mentioned earlier we will look at various optimization strategies.
|
||||||
\input{gfx/wifi_model_error_0_95.tex}
|
|
||||||
\input{gfx/wifi_model_error_95_100.tex}
|
|
||||||
\label{fig:wifiModelError}%
|
{\bf\noOptEmpiric{}} uses the same three parameters \mTXP,\mPLE,\mWAF for each \docAPshort{} in combination
|
||||||
\caption{%
|
with its position, which is well known from the flooprlan.
|
||||||
Comparison between different optimization strategies by examining the error (in \decibel) at each reference measurement.%
|
|
||||||
The higher the number of variable parameters, the better the model resembles real world conditions. %
|
{\bf\optParamsAllAP{}} is the same as above, except that the three parameters are optimized
|
||||||
}%
|
based on the reference measurements.
|
||||||
\end{figure}
|
|
||||||
|
{\bf\optParamsEachAP{}} optimizes the three parameters per \docAP{} instead of using the same
|
||||||
|
parameters for all.
|
||||||
|
|
||||||
|
{\bf\optParamsPosEachAP{}} does not need any prior knowledge and will optimize all six parameters
|
||||||
|
(3D position, \mTXP, \mPLE, \mWAF) based on the reference measurements.
|
||||||
|
|
||||||
|
{\bf\optPerFloor{}} and {\bf\optPerRegion{}} are just like \optParamsPosEachAP{} except that
|
||||||
|
there are several instances that are optimized only for one floor / region instead of the whole building.
|
||||||
|
|
||||||
|
\begin{figure}
|
||||||
|
\input{gfx/wifi_model_error_0_95.tex}
|
||||||
|
\input{gfx/wifi_model_error_95_100.tex}
|
||||||
|
\label{fig:wifiModelError}%
|
||||||
|
\caption{%
|
||||||
|
Comparison between different optimization strategies by examining the error (in \decibel) at each reference measurement.%
|
||||||
|
The higher the number of variable parameters, the better the model resembles real world conditions. %
|
||||||
|
}%
|
||||||
|
\end{figure}
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
@@ -215,10 +218,14 @@ kann man auch testen wenn man beim particle-filter das resampling ganz aus macht
|
|||||||
|
|
||||||
To reduce the amount such of misclassifications, where other locations within the building are
|
To reduce the amount such of misclassifications, where other locations within the building are
|
||||||
as likely as the pedestrians actual location, we examined various approaches.
|
as likely as the pedestrians actual location, we examined various approaches.
|
||||||
Unfortunately, none of which provided a viable enhancement under all conditions for the performed walks:
|
Unfortunately, none of which provided a viable enhancement under all conditions for the performed walks.
|
||||||
|
|
||||||
One possibility to dissolve an equal \docWIFI{}-likelihood between two (or more) locations within in the building
|
The misclassification-rate is determined by counting the amount of (random) locations within
|
||||||
is, to not only consider the \docAPshort{}s seen by the Smartphone, but also the \docAPshort{}s not seen
|
the building that produce a similar probability \refeq{eq:wifiProb} as the actual ground-truth
|
||||||
|
position.
|
||||||
|
|
||||||
|
One possibility to dissolve such an equal \docWIFI{}-likelihood between two (or more) locations is,
|
||||||
|
to not only consider the \docAPshort{}s seen by the Smartphone, but also the \docAPshort{}s not seen
|
||||||
by the Smartphone. This additional information can be used to rule out all locations where this
|
by the Smartphone. This additional information can be used to rule out all locations where this
|
||||||
\docAP{} should be received (high signal strength from the prediction model).
|
\docAP{} should be received (high signal strength from the prediction model).
|
||||||
% There might be an \docAP{} that should be visible at the other locations. However,
|
% There might be an \docAP{} that should be visible at the other locations. However,
|
||||||
@@ -234,20 +241,24 @@ kann man auch testen wenn man beim particle-filter das resampling ganz aus macht
|
|||||||
Furthermore, this requires the signal strength prediction model to be fairly accurate. Within our testing
|
Furthermore, this requires the signal strength prediction model to be fairly accurate. Within our testing
|
||||||
walks, several places are surrounded by concrete walls, which cause a harsh, local drop in signal strength.
|
walks, several places are surrounded by concrete walls, which cause a harsh, local drop in signal strength.
|
||||||
The models used within this work will not accurately predict the signal strength for such locations.
|
The models used within this work will not accurately predict the signal strength for such locations.
|
||||||
Including \docAPshort{}s unseen by the Smartphone thus often increases the estimation error instead
|
%%Including \docAPshort{}s unseen by the Smartphone thus often increases the estimation error instead
|
||||||
of fixing the multimodality.
|
%%of fixing the multimodality.
|
||||||
|
|
||||||
|
To sum up, while some situations, e.g. outdoors, could greatly be improved,
|
||||||
|
many other situations are deteriorated, especially when some transmitters are (temporarily)
|
||||||
|
attenuated by ambient conditions like concrete walls.
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
We therefore examined variations of the probability calculation from \refeq{eq:wifiProb}.
|
We therefore examined variations of the probability calculation from \refeq{eq:wifiProb}.
|
||||||
Removing the strongest/weakest \docAPshort{} from $\mRssiVecWiFi{}$ yielded similar results.
|
Removing the strongest/weakest \docAPshort{} from $\mRssiVec{}$ yielded similar results.
|
||||||
While some estimations were improved, the overall estimation error increased for our walks,
|
While some estimations were improved, the overall estimation error increased for our walks,
|
||||||
as there are many situations where only a handful \docAP{}s can be seen. Removing (valid)
|
as there are many situations where only a handful \docAP{}s can be seen. Removing (valid)
|
||||||
information will highly increase the error for such situations.
|
information will highly increase the error for such situations.
|
||||||
|
|
||||||
Using a more strict exponential distribution for the model vs. scan comparison in \refeq{eq:wifiProb}
|
Using a more strict exponential distribution for the model vs. scan comparison in \refeq{eq:wifiProb}
|
||||||
had a positive effect on the misclassification error for some of the walks, but slightly increased
|
had a positive effect on the misclassification error for some of the walks, but slightly increased
|
||||||
the estimation error (see figure \reffig{fig:normalVsExponential}) and thus produced negative side effects.
|
the estimation error (see figure \ref{fig:normalVsExponential}) and thus produced negative side effects.
|
||||||
|
|
||||||
\begin{figure}
|
\begin{figure}
|
||||||
\input{gfx/wifiCompare_normalVsExp_cross.tex}
|
\input{gfx/wifiCompare_normalVsExp_cross.tex}
|
||||||
@@ -261,38 +272,19 @@ kann man auch testen wenn man beim particle-filter das resampling ganz aus macht
|
|||||||
}
|
}
|
||||||
\end{figure}
|
\end{figure}
|
||||||
|
|
||||||
\todo{
|
|
||||||
wir wollen nicht, dass die position des ground-truths durch das wifi so wahrscheinlich wie möglich ist,
|
|
||||||
wir wollen dass die position des ground-truth einfach eine höhere wahrscheinlichkeit hat, als alle anderen punkte im gebäude
|
|
||||||
das pruefen wir ab
|
|
||||||
}
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
\todo{
|
\todo{
|
||||||
erkenntnisse:
|
erwähnen??? sigma je nach signalstärke anpassen bringt leider auch nichts. wenn man das aber macht,
|
||||||
|
dann: fuer grosse signalstaerken ein grosses sigma! andersrum gehts nach hinten los!
|
||||||
schlechte messwerte (niedrige RSSI) aus der messung ignorieren hilft nur sehr sehr bedingt.. eher im gegenteil. meist geht der fehler (stark) hoch
|
}
|
||||||
|
|
||||||
schlechteste messung weglassen ist auch schlecht
|
|
||||||
|
|
||||||
sigma je nach signalstärke anpassen bringt leider auch nichts. wenn man das aber macht,
|
|
||||||
dann: fuer grosse signalstaerken ein grosses sigma! andersrum gehts nach hinten los!
|
|
||||||
|
|
||||||
veto funktioniert auch nicht immer. es gibt stellen da ist ein AP wegen abschattung in der realität nicht sichtbar
|
|
||||||
das smpartphone sieht ihn deshalb nicht, im model ist er aber fälschlicherweise da deshalb falsches veto
|
|
||||||
oder das smartphone sieht einen AP wegen kollisionen nicht oder weil er durch den rücken stark verdeckt wird
|
|
||||||
|
|
||||||
es gibt einfach stellen an denen das wifi nicht eindeutig ist, die an anderen stellen quasi exakt genauso vorliegen
|
|
||||||
da laesst sich dann nicht viel machen
|
|
||||||
|
|
||||||
}
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
% -------------------------------- final system -------------------------------- %
|
||||||
|
|
||||||
% REAL WALKS
|
% REAL WALKS
|
||||||
\todo{obwohl das angepasste modell doch recht gut laeuft und der fehler recht klein wird, sind immernoch stellen dabei,
|
\todo{obwohl das angepasste modell doch recht gut laeuft und der fehler recht klein wird, sind immernoch stellen dabei,
|
||||||
wo es einfach nicht gut passt, unguenstige mehrdeutigkeiten vorliegen, oder regionen einfach nicht passen wie sie sollten.
|
wo es einfach nicht gut passt, unguenstige mehrdeutigkeiten vorliegen, oder regionen einfach nicht passen wie sie sollten.
|
||||||
@@ -301,7 +293,11 @@ kann man auch testen wenn man beim particle-filter das resampling ganz aus macht
|
|||||||
|
|
||||||
\todo{GPS ist leider kaum eine hilfe. entweder kein empfang wegen ueberdachung oder abschattung, oder
|
\todo{GPS ist leider kaum eine hilfe. entweder kein empfang wegen ueberdachung oder abschattung, oder
|
||||||
zu kurz draußen um einen guten gps-fix zu bekommen.}
|
zu kurz draußen um einen guten gps-fix zu bekommen.}
|
||||||
|
|
||||||
|
\todo{
|
||||||
|
walk1 hat eine issue kurz bevor man zur tuer zum hoersaalgebaude reingeht
|
||||||
|
je nach resampling killt dieser wlan error evtl alle partikel!
|
||||||
|
}
|
||||||
|
|
||||||
\todo{
|
\todo{
|
||||||
das bbox modell hat probleme an den uebergängen zwischen bboxes da dort teils starke spruenge sind
|
das bbox modell hat probleme an den uebergängen zwischen bboxes da dort teils starke spruenge sind
|
||||||
|
|||||||
@@ -1,4 +1,5 @@
|
|||||||
\section{Indoor Positioning System}
|
\section{Indoor Positioning System}
|
||||||
|
\label{sec:system}
|
||||||
|
|
||||||
Our smartphone-based indoor localization system estimates the current location and heading
|
Our smartphone-based indoor localization system estimates the current location and heading
|
||||||
using recursive density estimation.
|
using recursive density estimation.
|
||||||
|
|||||||
@@ -1,4 +1,5 @@
|
|||||||
\section{WiFi Optimization}
|
\section{WiFi Location Estimation}
|
||||||
|
\label{sec:optimization}
|
||||||
|
|
||||||
The WiFi sensor infers the pedestrian's current location based on a comparison between live measurements
|
The WiFi sensor infers the pedestrian's current location based on a comparison between live measurements
|
||||||
(the smartphone continuously scans for nearby \docAP{}s) and reference measurements / predictions
|
(the smartphone continuously scans for nearby \docAP{}s) and reference measurements / predictions
|
||||||
@@ -146,17 +147,46 @@
|
|||||||
|
|
||||||
\subsection{Modified Signal Strength Model}
|
\subsection{Modified Signal Strength Model}
|
||||||
|
|
||||||
\todo{nicht: during initial eval, sondern gleich sagen, dass die vermutung nahe liegt, dass das modell
|
%\todo{nicht: during initial eval, sondern gleich sagen, dass die vermutung nahe liegt, dass das modell
|
||||||
nicht gut klappen wird, weil waende und unser metall-glas nicht beruecksichtigt werden. deshalb
|
%nicht gut klappen wird, weil waende und unser metall-glas nicht beruecksichtigt werden. deshalb
|
||||||
versuchen wir ein anderes modell das immernoch live arbeiten kann}
|
%versuchen wir ein anderes modell das immernoch live arbeiten kann}
|
||||||
During the initial eval, some issues were discovered. While aforementioned optimization was able to
|
|
||||||
reduce the error between reference measurements and model estimations to \SI{50}{\percent},
|
%During the initial eval, some issues were discovered. While aforementioned optimization was able to
|
||||||
the position estimation \ref{eq:wifiProb} did not benefit from improved model parameters.
|
%reduce the error between reference measurements and model estimations to \SI{50}{\percent},
|
||||||
To the contrary, there were several situations throughout the testing walks, where
|
%the position estimation \ref{eq:wifiProb} did not benefit from improved model parameters.
|
||||||
the inferred location was more erroneous than before.
|
%To the contrary, there were several situations throughout the testing walks, where
|
||||||
|
%the inferred location was more erroneous than before.
|
||||||
|
|
||||||
|
As the used model does not consider walls, it is expected to provide erroneous values
|
||||||
|
for regions that are heavily attenuated by e.g. concrete or metallised glass.
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
\subsection{\docWIFI{} quality factor}
|
||||||
|
|
||||||
|
\todo{wifi quality factor??}
|
||||||
|
\todo{formel für toni}
|
||||||
|
|
||||||
|
\begin{equation}
|
||||||
|
\newcommand{\leMin}{l_\text{min}}
|
||||||
|
\newcommand{\leMax}{l_\text{max}}
|
||||||
|
q(\mRssiVec) =
|
||||||
|
\max(0,
|
||||||
|
\min(
|
||||||
|
\frac{
|
||||||
|
\bar\mRssi - \leMin
|
||||||
|
}{
|
||||||
|
\leMax - \leMin
|
||||||
|
},
|
||||||
|
1
|
||||||
|
)
|
||||||
|
)
|
||||||
|
\label{eq:wifiQuality}
|
||||||
|
\end{equation}
|
||||||
|
|
||||||
|
|
||||||
\subsection {VAP grouping}
|
\subsection {VAP grouping}
|
||||||
|
\label{sec:vap}
|
||||||
|
|
||||||
Assuming normal conditions, the received signal strength at one location will (strongly) vary
|
Assuming normal conditions, the received signal strength at one location will (strongly) vary
|
||||||
due to environmental conditions like temperature, humidity, open/closed doors, RF interference.
|
due to environmental conditions like temperature, humidity, open/closed doors, RF interference.
|
||||||
|
|||||||
Reference in New Issue
Block a user