Mobile Phone Localization via Ambience Fingerprinting Martin Azizyan Duke University Ionut Constandache Duke University Romit Roy Choudhury Duke University Abstract Mobile computing ID: 591464
Download Presentation The PPT/PDF document "SurroundSense" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
SurroundSense:Mobile Phone Localizationvia Ambience Fingerprinting
Martin
Azizyan
Duke University
Ionut
Constandache
Duke University
Romit
Roy Choudhury Duke UniversitySlide2
AbstractMobile computing applications center around user’s locationTerm- Physical Location
(
coordinates Latitude and Longitude)
Term- Logical Location
(
like Starbucks or McDonalds)
Lots of Research available for physical
location
Few attempts recognizing logical
locationSlide3
AbstractAmbient sound, light, and color from phone’s camera and
microphone
A
ccelerometers for user-motion
Adjacent
stores can be separated
logically
They
propose
SurroundSense
, a mobile phone based system
explores ambience fingerprinting
51
different stores
- average
accuracy of 87% when all sensing modalities are
employedSlide4
IntroductionStarbucks -coffee machines and microwaves
Restaurants
-forks
and spoons
clinking
Target – red colors
Panera Breads- yello
w colors
Floors
with
carpets
, ceramic
tiles, or wooden
strips
Bars -dim
yellow lights
BlockBuster
-
bright white
lightSlide5
IntroductionWal-Mart - walking up and down aisles
Barnes
and Noble
-relaxed
stroll with long
pauses
R
estaurants - short
queuing
followed by
a long duration of
sitting
May not
be unique based on any one attribute, the
combination of
all
exhibits diversitySlide6
SurroundSense ArchitectureSlide7
SurroundSense ArchitectureMobile phone user visits
an
unknown store
.
The
phone senses the
ambience
V
alues forwarded
to the fingerprinting
factory
Types
of
sensor data
identified
P
hone’s
(
GSM-based) physical
coordinates
Geographical database
Fingerprint database and Fingerprint matching
The matching module
-best match for test fingerprintSlide8
System DesignFingerprinting SoundFingerprinting Motion using AccelerometersFingerprinting Color/Light using Cameras
Fingerprinting Wi-Fi
Fingerprint MatchingSlide9
Fingerprinting SoundRecorded ambient sound for one
Time
domain, a
simple fingerprinting scheme based
on
signal amplitude
Acoustic
fingerprints
- computed
the
pair-wise distances
.
Use
sound only as a
filter
The
output
is
fed to the accelerometer
filterSlide10
Sound fingerprints from 3 adjacent storesSlide11
Fingerprinting Motion using AccelerometersHuman movements in a location
Restaurants
-stationary
for long durations
G
rocery
store
- more
mobile
Accelerometer readings -
stationary
vs
in
motion
Vector
machines (SVM),
a popular
data classification
tool
User
movement is prone to
fluctuation
Clothing
store
-
browse long
time
or purchase in haste
A
ccelerometers
as
a filtering mechanism
tooSlide12
Sample Accelerometers TracesSlide13
Fingerprinting Color/Light using CamerasThe wall and floor
colors contribute to theme
Use automatically-taken
phone
pictures
Only floor-facing
pictures are
used
Color/light extraction
Why picture of the floor?
Colors
of carpets
, tiles
, marble, and wooden
floors
HSL - hue-saturation-lightness
Clusters
of color
color-light fingerprintSlide14
Color/light fingerprint in the HSL spaceSlide15
Fingerprinting Wi-FiWiFi fingerprinting no good for
logical places
WiFi
based fingerprinting
-fifth
sensor
MAC
addresses of
visible APs
MAC
addresses
recorded every
5
seconds
Computing
the
fraction of times each unique MAC address
was seen
over all
recordings
A
tuple of fractions
forms
the
WiFi
fingerprint of that
placeSlide16
Fingerprint MatchingSurroundSense uses 4 filtering/matching The (WiFi
, sound, and accelerometer) filters are applied first
Candidate
set
fed
to the color/light-based matching
scheme
Use
the color/light based matching scheme
last
The
final
output is an ordered list of candidates – the top
ranked candidate
is declared to be the location of the
phoneSlide17
ProtoType ImplementationClient and ServerPopulating the Fingerprint Database
Special Note:
SurroundSense
was implemented on Nokia N95
using
Python
platform
.
The
server
-MATLAB
,
Python
code,
data mining toolsSlide18
Client and ServerSensor runs on threads and execute
API
calls
The
accelerometer
samples - 4
readings per second
.
The
audio sampling rate is 8 kHz.
Pictures
are taken every
5
seconds
A
meta file
-stores date
,
time
,
GSM,
camera
mode
The server
- several modules.
A
Data
Manager
formats raw data appropriately.
The
formatted data
sent to
Fingerprinting
Factory
A MATLAB
/ Python
based Filtering/Matching Module
-computes
the top-ranked match.Slide19
Populating the Fingerprint DatabaseHow did they build a fingerprint database?
46
business locations
5
locations in
India
Students
visited 51
stores
Stores visited
multiple times
Design
location
labeling
games. The
person with a best match may win a prize.
More people
play
larger
fingerprint
databaseSlide20
EvaluationPartially Controlled ExperimentationPerformance Pre-Cluster Accuracy
Per-Shop Accuracy
Per-User Accuracy
Per-Sensor AccuracySlide21
Partially Controlled ExperimentationNot performed with a real user base Mobile
phones
in our hand (and not in our pockets
)
Phones took pictures
for color and light fingerprinting.
In uncontrolled environments
, phones
in pocket
New wearable mobile
phones
Wrist
watches and
N
ecklacesSlide22
Mimicking Customer BehaviorGroups of 2 people Went
to different stores
-time-separated
Fingerprinted
every store in
cluster
Behave
like normal
customers
Purchase
coffee
and food
Mimic
the movement of another
customers
Atypical behavior -picking
up pre-ordered food
shopping
very quickly
. Slide23
Performance: Per-Cluster AccuracyEvaluate 4 modes1.
WiFi
-only
2
.
Sound, Accelerometer, Light and Color
3
.
Sound, Accelerometer
4
.
SurroundSense
(SS)
combined all
modes
of ambience
fingerprinting
SurroundSense
average
accuracy
of 87%
All the sensors
90%...Slide24
Performance: Per-Shop Accuracy47% of the shops can be localized perfectly using SurroundSense.
WiFi
displays
bimodal
behavior –
it’s
either
high
accuracy, or
seriously suffers
Clearly, the combination of multi-modal
fingerprinting is bestSlide25
Performance: Per-User AccuracyUser assigned to a random set of stores
Report
the average
accuracy
SurroundSense
users
achieve between 73% and 75%
accuracy
The accuracy grows
to an average of 83% or more for 80% of the
users
The median accuracy is around 88%, while 10% users experience 96% accuracy or
moreSlide26
Performance: Per-Sensor Accuracyhand-picked 6 examples -merits and demerits of each
sensor
Whenever
the accelerometer
is
used-accuracy
is always 100
%
Only
the camera,
-100
%
accuracy in
this
location
Only
color gives average accuracy of 91%. When
sound is
added,
-66%
If the correct
location is filtered out, the final match
incorrectSlide27
Limitations and Future WorkEnergy ConsiderationsEnergy efficient localization
and
sensing
Simple
sensing mechanisms
– when outdoors
Variation in GSM signal strengths
T
emperature
sensing
An
accelerometer trace requires
time
Faster
methods of
localization without
compromising
accuracySlide28
ConclusionLogical location, vs physical coordinates The
main idea
-ambient
sound, light, color, RF,
luser
movement
Fingerprint identifies user’s location
SurroundSense
not a stand-alone technique- use with GSM location
SurroundSense
step
towards
indoor localization
Further
research
-
better
energy management
SurroundSense
a viable solution of the
futureSlide29
Questions/Comments?