View
217
Download
1
Category
Preview:
Citation preview
The Horus WLAN Location Determination System
Moustafa Youssef and Ashok Agrawala
University of Maryland, College Park
June, 2005H
O
R
U
S
H
O
R
U
S
H
O
R
U
S
H
O
R
U
S
2June, 2005 Horus © 2005, Moustafa Youssef
Location Determination Technologies
GPS Cellular-based Ultrasonic-based: Active Bat Infrared-based: Active Badge Computer vision: Easy Living Physical proximity: Smart Floor Not suitable for indoor
– Does not work– Require specialized hardware– Scalability
H
O
R
U
S
H
O
R
U
S
3June, 2005 Horus © 2005, Moustafa Youssef
WLAN Location Determination
Triangulate user location
– Reference point
– Quantity proportional to distance
WLAN
– Access points
– Signal strength= f(distance)
Software based
H
O
R
U
S
H
O
R
U
S
4June, 2005 Horus © 2005, Moustafa Youssef
Roadmap
Motivation
Location determination technologies
Introduction
Noisy wireless channel
Horus components
Performance evaluation
Conclusions and future work
H
O
R
U
S
H
O
R
U
S
5June, 2005 Horus © 2005, Moustafa Youssef
WLAN Location Determination (Cont’d)
Signal strength= f(distance) Does not follow free space loss Use lookup table Radio map Radio Map: signal strength characteristics at selected
locations
H
O
R
U
S
H
O
R
U
S
6June, 2005 Horus © 2005, Moustafa Youssef
WLAN Location Determination Taxonomy
WLAN Location Determination Systems
Ad-hoc Mode Infrastructure Mode
Cell of Origin Time of ArrivalSignal Strength
Model-based Radio-map Based
Radar Horus
Daedalus PinPoint
[Lundberg02]
Classification
Example
Deterministic ProbabilisticWheremops
H
O
R
U
S
H
O
R
U
S
7June, 2005 Horus © 2005, Moustafa Youssef
WLAN Location Determination (Cont’d)
Offline phase– Build radio map– Radar system: average signal strength
Online phase– Get user location– Nearest location in signal strength space (Euclidian
distance)
[-53, -56]
[-50, -60]
[-58, -68]
5
13
H
O
R
U
S
H
O
R
U
S
8June, 2005 Horus © 2005, Moustafa Youssef
Horus Goals
High accuracy
– Wider range of applications
Energy efficiency
– Energy constrained devices
Scalability
– Number of supported users
– Coverage area
H
O
R
U
S
H
O
R
U
S
9June, 2005 Horus © 2005, Moustafa Youssef
Roadio-map
Motivation
Location determination technologies
Introduction
Noisy wireless channel
Horus components
Performance evaluation
Conclusions
H
O
R
U
S
H
O
R
U
S
10June, 2005 Horus © 2005, Moustafa Youssef
Sampling Process
Active scanning– Send a probe
request– Receive a probe
responseChannel 2
Channel 1
...
1. Probe Request
2. Probe Response
3. Probe Request
4. Probe R
esponse
Channel n
2n-1
. P
robe R
equest
2n. P
robe R
esp
onse
H
O
R
U
S
H
O
R
U
S
11June, 2005 Horus © 2005, Moustafa Youssef
Signal Strength Characteristics
Temporal variations
– One access point
– Multiple access points
Spatial variations
– Large scale
– Small scale
H
O
R
U
S
H
O
R
U
S
12June, 2005 Horus © 2005, Moustafa Youssef
Temporal Variations:One Access Point
Environment changes
Using average only leads to loss of information
H
O
R
U
S
H
O
R
U
S
13June, 2005 Horus © 2005, Moustafa Youssef
Temporal Variations:Multiple Access Points
0
50
100
150
200
250
300
-95 -85 -75 -65 -55
Average Signal Strength (dBm)
Nu
mb
er
of
Sam
ple
s
Co
llecte
dReceiver Sensit ivity
Number of access points changes over time
Choose the strongest access points
H
O
R
U
S
H
O
R
U
S
14June, 2005 Horus © 2005, Moustafa Youssef
Temporal Variations:Correlation
Independence assumption is wrong
H
O
R
U
S
H
O
R
U
S
15June, 2005 Horus © 2005, Moustafa Youssef
Spatial Variations: Large-Scale
-65
-60
-55
-50
-45
-40
-35
-30
0 5 10 15 20 25 30 35 40 45 50 55
Distance (feet)
Sig
na
l S
tre
ng
th
(db
m)
Desirable !
H
O
R
U
S
H
O
R
U
S
16June, 2005 Horus © 2005, Moustafa Youssef
Spatial Variations: Small-Scale
Multipath effect
H
O
R
U
S
H
O
R
U
S
17June, 2005 Horus © 2005, Moustafa Youssef
Roadio-map
Motivation
Goals
Introduction
Noisy wireless channel
Horus components
Performance evaluation
Conclusions
H
O
R
U
S
H
O
R
U
S
18June, 2005 Horus © 2005, Moustafa Youssef
Horus Components
Basic algorithm
Correlation handler
Continuous space estimator
Small-scale compensator
Locations clustering
H
O
R
U
S
H
O
R
U
S
19June, 2005 Horus © 2005, Moustafa Youssef
Offline phase
– Radio map: signal strength histograms
Online phase
– Bayesian based inference
Basic Algorithm
H
O
R
U
S
H
O
R
U
S
20June, 2005 Horus © 2005, Moustafa Youssef
Basic Algorithm:Example
(xi, yi)
(x, y)
-40 -60 -80
-40 -60 -80
[-53]
P(-53/L1)=0.55
P(-53/L2)=0.08
H
O
R
U
S
H
O
R
U
S
21June, 2005 Horus © 2005, Moustafa Youssef
Using Multiple Samples
Need to average multiple samples to increase accuracy
Independence assumption is wrong
H
O
R
U
S
H
O
R
U
S
22June, 2005 Horus © 2005, Moustafa Youssef
Autoregressive model
– Estimate correlation degree
– Estimate distribution of the average of n correlated samples
Correlation Handler
H
O
R
U
S
H
O
R
U
S
23June, 2005 Horus © 2005, Moustafa Youssef
Correlation Handler:Var(A)/Var(s)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.2 0.4 0.6 0.8 1a
Va
r(A
)/V
ar(
s)
0 1 2 3 4 5 6 7 8 9 10
Independence assumption underestimates true variance
H
O
R
U
S
H
O
R
U
S
24June, 2005 Horus © 2005, Moustafa Youssef
Enhance the discrete radio map space estimator
Two techniques
– Center of mass of the top ranked locations
– Time averaging window
Continuous Space Estimator
H
O
R
U
S
H
O
R
U
S
25June, 2005 Horus © 2005, Moustafa Youssef
Small-scale Compensator
Perturbation Technique
Detect small-scale variations
– Using previous user location
Perturb signal strength vector
– (s1, s2, …, sn) (s1d1, s2d2, …, sndn)
– Typically, n=3-4
di is chosen relative to the received signalstrength
H
O
R
U
S
H
O
R
U
S
26June, 2005 Horus © 2005, Moustafa Youssef
Reduce computational requirements
Use access points that cover each location
Use the q strongest access points
Locations Clustering
0
50
100
150
200
250
300
-95 -85 -75 -65 -55
Average Signal Strength (dBm)
Nu
mb
er
of
Sam
ple
s
Co
llecte
d
Receiver Sensit ivity
H
O
R
U
S
H
O
R
U
S
27June, 2005 Horus © 2005, Moustafa Youssef
Horus Components
Discrete-Space
Estimator
Continuous-Space
Estimator
Small-Scale
Compensator
Correlation
HandlerClustering
Correlation
Modeler
Radio Map
Builder
Radio
Map
and
clusters
Ho
rus S
yste
m C
om
po
ne
nts
Location API
Applications
Signal Strength Acquisition API
Estimated Location
Device Driver
(MAC, Signal Strength)(-50,-67,-80)
(-45,-63,-63)
H
O
R
U
S
H
O
R
U
S
28June, 2005 Horus © 2005, Moustafa Youssef
Roadio-map
Motivation
Location Determination technologies
Introduction
Noisy wireless channel
Horus components
Performance evaluation
Conclusions and future work
H
O
R
U
S
H
O
R
U
S
29June, 2005 Horus © 2005, Moustafa Youssef
Testbeds
A.V. William’s
– 4th floor, AVW
– 224 feet by 85.1 feet
– UMD net (Cisco APs)
– 21 APs (6 on avg.)
– 172 locations
– 5 feet apart
– Windows XP Prof.
FLA
– 3rd floor, 8400 Baltimore Ave
– 39 feet by 118 feet
– LinkSys/Cisco APs
– 6 APs (4 on avg.)
– 110 locations
– 7 feet apart
– Linux (kernel 2.5.7)
Orinoco/Compaq cards
H
O
R
U
S
H
O
R
U
S
30June, 2005 Horus © 2005, Moustafa Youssef
Horus-Radar Comparison
0
500
1000
1500
2000
2500
3000
3500
4000
Horus RadarAv
g. N
um
. o
f O
pe
r. p
er L
oc
. E
st.
Median Avg Stdev Max
Horus (all components) 1.28 1.38 0.95 4
Horus (basic) 1.6 2.16 2.09 18.08
Radar 9.74 13.15 10.71 57.67
H
O
R
U
S
H
O
R
U
S
31June, 2005 Horus © 2005, Moustafa Youssef
Radar with Horus Techniques
Average distance error enhanced by more than 58%
Worst case error decreased by more than 76%
H
O
R
U
S
H
O
R
U
S
32June, 2005 Horus © 2005, Moustafa Youssef
Roadio-map
Motivation
Location Determination technologies
Introduction
Noisy wireless channel
Horus components
Performance evaluation
Conclusions
H
O
R
U
S
H
O
R
U
S
33June, 2005 Horus © 2005, Moustafa Youssef
Conclusions
The Horus system achieves its goals
High accuracy– through different modules
Low computational requirements– through the use of clustering techniques
Scalability in terms of the coverage area– through the use of clustering techniques
Scalability in terms of the number of users– through the distributed implementation
Modules can be applied to other WLAN location determination systems
H
O
R
U
S
H
O
R
U
S
34June, 2005 Horus © 2005, Moustafa Youssef
Other Horus Related
Invention of the year award (UMD 2004) 3 Patents pending Licensed by Fujitsu Cited in
– New York Times– Washington Times
Software– Drivers: mwvlan, mwavelan, morinoco– MAPI– http://www.cs.umd.edu/~moustafa
H
O
R
U
S
H
O
R
U
S
35June, 2005 Horus © 2005, Moustafa Youssef
For More Information
Overall system [MobiSys05]
Basic algorithm [Percom03]
Locations clustering [Percom03]
Small-scale compensator [WCNC03]
Optimality Analysis [CNDS04]
Correlation handler [InfoCom04]
Continuous space estimator [ICCCN04]
User profile [IJMS05]
Drivers/API’s
www.cs.umd.edu/users/moustafa
Recommended