authentication using pulse-responce biometrics€¦ · pulse-responce biometrics kasper b....
TRANSCRIPT
February 25, 2014 (slide 1)
Authentication UsingPulse-Responce Biometrics
Kasper B. Rasmussen1 Marc Roeschlin2
Ivan Martinovic1 Gene Tsudik3
1University of Oxford 2ETH Zurich 3UC Irvine
February 25, 2014
February 25, 2014 (slide 2)
Outline
1 Background on Biometrics
2 Pulse-Response
3 Security Application
4 Experimental Results
February 25, 2014 (slide 3)
Biometrics: A Definition
BiometricsMechanism to identify people by characteristics ortraits.
February 25, 2014 (slide 4)
Biometrics
BehavioralIncludes: keystroke timings, speechpattern analysis, gait recognition,and analysis of stylus pressure,acceleration and shape inhand-writing
PhysiologicalIncludes: fingerprints, handgeometry, facial recognition, speechanalysis, and iris/retina scans
February 25, 2014 (slide 5)
Pulse-Response Biometric
Pulse signal applied to thepalm of one hand.
The biometric is captured bymeasuring the response inthe user’s hand.
February 25, 2014 (slide 6)
User Safety
Voltage (V) 1 1.5
Max Current (mA) 0.1 500+
Exposure 100ns ∼500ms
February 25, 2014 (slide 7)
Design Goals
Universal: Usable by everyone that needs touse it.
Unique: within the target population.
Permanent: Consistent over the time (stable).
Unobtrusive: Does not disrupt the user’s workflow.
Difficult to circumvent: Unable to modify thebiometric or impersonate others.
February 25, 2014 (slide 8)
Biometrics in Security
IdentificationObtain the identity of a user.
vs.
AuthenticationConfirm the identity of a user.
Continuous AuthenticationContinuously confirm the identity of a user.
February 25, 2014 (slide 9)
Biometrics in Security
IdentificationObtain the identity of a user.
vs.
AuthenticationConfirm the identity of a user.
Continuous AuthenticationContinuously confirm the identity of a user.
February 25, 2014 (slide 10)
Hardening PIN Entry Systems
February 25, 2014 (slide 11)
ATM Decision Flowchart
End
Start Is PINCorrect?
Accept! Reject!Does
pulse-responsematch?
No
No
Yes
Yes
February 25, 2014 (slide 12)
Experimental Setup
February 25, 2014 (slide 13)
Signals
0.0
0.5
1.0
0 200 400 600 800
Time [ns]
Sig
nal m
agni
tude
[Vol
t]
Input pulse
Measured pulse
0
100
200
300
400
500
0 25 50 75 100
Frequency bins
Spe
ctra
l den
sity
February 25, 2014 (slide 14)
Classification
February 25, 2014 (slide 15)
Authentication Classifier
Over Time
Aiden Ethan
Jacob Liam
Mason
0%
25%
50%
75%
100%
0%
25%
50%
75%
100%
0%
25%
50%
75%
100%
90 92 94 96 98 100Threshold [%]
Sen
sitiv
ity (
TP
R)
February 25, 2014 (slide 16)
Single SessionCharles David
Ethan Jackson
Liam Lucas
Mason Noah
Richard Sophia
0%
25%
50%
75%
100%
0%
25%
50%
75%
100%
0%
25%
50%
75%
100%
0%
25%
50%
75%
100%
0%
25%
50%
75%
100%
90 92 94 96 98 100 90 92 94 96 98 100Threshold [%]
Sen
sitiv
ity (
TP
R)
February 25, 2014 (slide 17)
Identification Classifier
Over time Single data set
0%
25%
50%
75%
100%
Aide
n
Etha
n
Jaco
b
Liam
Mas
on
Cha
rles
Dav
id
Etha
nJa
ckso
n
Liam
Luca
s
Mas
on
Noa
hR
icha
rd
Soph
ia
Sen
sitiv
ity (
TP
R)
February 25, 2014 (slide 18)
Future Work
Prototype
Build PIN entry prototype.
Gather experience on acquisition time, etc.
Gather data.
Acquisition Signal
Higher bandwidth
No signal
Effects of stress, blood sugar levels, etc.
Test impersonation strategies.
February 25, 2014 (slide 19)
Conclusion
A new biometric based on Pulse-Response.
A scenario where Pulse-Response can be easilyintegrated.
Fantastically promising results. Very highdegree of uniqueness and remarkably stableover time.
Thank you for your attention. Questions?
February 25, 2014 (slide 20)
February 25, 2014 (slide 21)
Conclusion – Questions?
A new biometric based on Pulse-Response.
A scenario where Pulse-Response can be easilyintegrated.
Fantastically promising results. Very highdegree of uniqueness and remarkably stableover time.
Thank you for your attention. Questions?