On the Security of Carrier Phase-based Ranging Hildur Olafsdottir, - - PowerPoint PPT Presentation
On the Security of Carrier Phase-based Ranging Hildur Olafsdottir, - - PowerPoint PPT Presentation
On the Security of Carrier Phase-based Ranging Hildur Olafsdottir, Aanjhan Ranganathan, Srdjan Capkun Importance of Secure Proximity Verification Aanjhan Ranganathan On the Security of Carrier Phase-based Ranging 2 Importance of Secure
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Importance of Secure Proximity Verification
2
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Importance of Secure Proximity Verification
2
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Importance of Secure Proximity Verification
2
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Importance of Secure Proximity Verification
2
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Importance of Secure Proximity Verification
2
Distance is determined using a variety of methods (e.g., based on received signal strength, time of flight etc.)
Physical-layer Techniques for Secure Proximity Verification & Localization
Estimating Distance
3
d
Received Signal Strength (RSS)
ttof tp
d = c * (ttof - tp) / 2
Time of Flight (ToF)
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Attacks on Proximity Systems
4
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Attacks on Proximity Systems
4
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Attacks on Proximity Systems
4
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Attacks on Proximity Systems
4
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Attacks on Proximity Systems
4
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Attacks on Proximity Systems
4
No knowledge of the data exchanged is required! Independent of cryptographic primitives
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Multi-carrier Phase Ranging
5
- Cost optimised solution for proximity-based applications
- Low-hardware complexity, low-power consumption, high precision
- Compliant with prominent standards such as WiFi, ZigBee
- Localization schemes* leveraging signal phase information are
increasingly becoming popular
- 802.11, NB-IoT, LoRa, 5G networks
* Vasisht, Deepak, Swarun Kumar, and Dina Katabi. "Decimeter-Level Localization with a Single WiFi Access Point." NSDI 2016. * Xiong, J., Sundaresan, K., Jamieson, K. “Tonetrack: Leveraging frequency-agile radios for time-based indoor wireless localization.” MobiCom 2015. * Exel, R. “Carrier-based ranging in ieee 802.11 wireless local area networks.” IEEE Wireless Communications and Networking Conference (WCNC) 2013.
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Motivation
6
- Implications of distance modification attacks are significant
- loss of property to even human life (e.g., IMD access control)
- Security of multi-carrier phase ranging has not be analysed yet.
- number of prior works on other prominent ranging systems*
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Contributions
7
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Contributions
7
- Investigate the vulnerability of carrier phase-based ranging system to
distance modification attacks
- focus on distance reduction attacks
- no knowledge of the implemented cryptographic primitive (if any) required
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Contributions
7
- Investigate the vulnerability of carrier phase-based ranging system to
distance modification attacks
- focus on distance reduction attacks
- no knowledge of the implemented cryptographic primitive (if any) required
- Three different attack realisations (varying attacker complexity)
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Contributions
7
- Investigate the vulnerability of carrier phase-based ranging system to
distance modification attacks
- focus on distance reduction attacks
- no knowledge of the implemented cryptographic primitive (if any) required
- Three different attack realisations (varying attacker complexity)
- Experiments show that it is possible to reduce the estimated distance to
less than 3 m even though the devices were more than 50 m apart
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Contributions
7
- Investigate the vulnerability of carrier phase-based ranging system to
distance modification attacks
- focus on distance reduction attacks
- no knowledge of the implemented cryptographic primitive (if any) required
- Three different attack realisations (varying attacker complexity)
- Experiments show that it is possible to reduce the estimated distance to
less than 3 m even though the devices were more than 50 m apart
50 m
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Contributions
7
- Investigate the vulnerability of carrier phase-based ranging system to
distance modification attacks
- focus on distance reduction attacks
- no knowledge of the implemented cryptographic primitive (if any) required
- Three different attack realisations (varying attacker complexity)
- Experiments show that it is possible to reduce the estimated distance to
less than 3 m even though the devices were more than 50 m apart
50 m < 3 m
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Phase-based Ranging
8
d
θ Δ θ Δ
7
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Phase-based Ranging
8
θ
d = c 2 · f · ( θ 2π + n)
d
θ Δ θ Δ
7
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Phase-based Ranging
8
θ
d = c 2 · f · ( θ 2π + n)
d
θ Δ θ Δ
7
ambiguity!
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Phase-based Ranging
8
θ
d = c 2 · f · ( θ 2π + n)
f1 f2 θ1 θ2
d = c 4π · θ2 − θ1 f2 − f1
d
θ Δ θ Δ
7
ambiguity!
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Distance Decreasing Relay Attacks
9
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Distance Decreasing Relay Attacks
9
System Assumption
- Two entities, verifier (e.g., car) and a prover (e.g., key) estimate distance
using multicarrier phase ranging technology
- Verifier and prover implement some form of cryptographic authentication
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Distance Decreasing Relay Attacks
9
System Assumption
- Two entities, verifier (e.g., car) and a prover (e.g., key) estimate distance
using multicarrier phase ranging technology
- Verifier and prover implement some form of cryptographic authentication
Attacker Model
- Verifier and prover are trusted and assumed to be honest
- External attacker tries to reduce the estimated distance between a honest
prover and verifier
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan 10
Phase slope Rollover Attack
d = c 4π · θ2 − θ1 f2 − f1
Distance
f1 f2 θ1 θ2
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan 10
Phase slope Rollover Attack
d = c 4π · θ2 − θ1 f2 − f1
Distance
dmax =
c 4π · ∆θmax ∆f
Maximum measurable distance
f1 f2 θ1 θ2
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan 10
Phase slope Rollover Attack
0 to 2π
d = c 4π · θ2 − θ1 f2 − f1
Distance
dmax =
c 4π · ∆θmax ∆f
Maximum measurable distance
f1 f2 θ1 θ2
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan 10
Phase slope Rollover Attack
dmax = c
2 · 1 ∆f
0 to 2π
d = c 4π · θ2 − θ1 f2 − f1
Distance
dmax =
c 4π · ∆θmax ∆f
Maximum measurable distance
f1 f2 θ1 θ2
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan 10
Phase slope Rollover Attack
dmax = c
2 · 1 ∆f
0 to 2π
∆f
e.g., = 2 MHz, then dmax = 75 m after which distance rollsover back to 0
d = c 4π · θ2 − θ1 f2 − f1
Distance
dmax =
c 4π · ∆θmax ∆f
Maximum measurable distance
f1 f2 θ1 θ2
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan 10
Phase slope Rollover Attack
dmax = c
2 · 1 ∆f
0 to 2π
∆f
e.g., = 2 MHz, then dmax = 75 m after which distance rollsover back to 0
d = c 4π · θ2 − θ1 f2 − f1
Distance
dmax =
c 4π · ∆θmax ∆f
Maximum measurable distance
Attacker leverages this maximum measurable distance property to reduce the estimated distance
f1 f2 θ1 θ2
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan 11
V P θ t Δt θ Δt
Phase slope Rollover Attack
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan 11
V P θ t Δt θ Δt
Phase slope Rollover Attack
0.2 0.4 0.6 0.8 Delay [7s] 20 40 60 80
Measured Distance [m]
∆f = 2 MHz, then dmax = 75 m
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan 11
V P θ t Δt θ Δt
Phase slope Rollover Attack
0.2 0.4 0.6 0.8 Delay [7s] 20 40 60 80
Measured Distance [m]
∆f = 2 MHz, then dmax = 75 m
500 ns
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Practical Demonstration of the Attack
12
~50m
dap 3 4 2 1
Experimental Setup
- Two Atmel AT86RF233 multicarrier phase ranging
- devices. (1) prover, (3) verifier
- Attacker hardware (2): USRP and two directional
antennas
- delay logic implemented directly on the FPGA
- Laptop (4) records the verifier’s distance estimates
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan 13 13
relay switched on
Practical Demonstration of the Attack
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
RF Cycle Slip Attack
14
V P θ t Δt1 θ Δt2
The attacker intercepts the prover’s signal and delays each frequency individually. Requires nanosecond level control of the relay signal Limitation: Attacker needs very high sampling rate. May alternatively use analog delay lines.
d = c 4π · θ2 − θ1 f2 − f1
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
RF Cycle Slip Attack
14
V P θ t Δt1 θ Δt2
The attacker intercepts the prover’s signal and delays each frequency individually. Requires nanosecond level control of the relay signal Limitation: Attacker needs very high sampling rate. May alternatively use analog delay lines.
d = c 4π · θ2 − θ1 f2 − f1
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
On-the-fly Phase Manipulation
15
θ t V P θap θA θA-θap
sa(t) sif(t) sp(t)
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
On-the-fly Phase Manipulation
15
θ t V P θap θA θA-θap
sA(t) = sif(t) ⊗ sP (t)
= cos(4πft + θA) ⊗ cos(2πft + θap)
sa(t) sif(t) sp(t)
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
On-the-fly Phase Manipulation
15
θ t V P θap θA θA-θap
sA(t) = sif(t) ⊗ sP (t)
= cos(4πft + θA) ⊗ cos(2πft + θap)
sa(t) sif(t) sp(t)
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
On-the-fly Phase Manipulation
15
θ t V P θap θA θA-θap
sA(t) = sif(t) ⊗ sP (t)
= cos(4πft + θA) ⊗ cos(2πft + θap)
LP
=
1 2 cos
⇣ 2πft + θA − θap ⌘
sa(t) sif(t) sp(t)
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
On-the-fly Phase Manipulation
15
θ t V P θap θA θA-θap
sA(t) = sif(t) ⊗ sP (t)
= cos(4πft + θA) ⊗ cos(2πft + θap)
LP
=
1 2 cos
⇣ 2πft + θA − θap ⌘
sa(t) sif(t) sp(t)
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
On-the-fly Phase Manipulation
15
θ t V P θap θA θA-θap
sA(t) = sif(t) ⊗ sP (t)
= cos(4πft + θA) ⊗ cos(2πft + θap)
LP
=
1 2 cos
⇣ 2πft + θA − θap ⌘
sa(t) sif(t) sp(t) Almost real time except for the delay from the mixer, low pass filter
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Possible Countermeasures
16
- Frequency Hopping
- nly valid if attacker is incapable of processing a wide band of frequencies
- simultaneously. Very weak assumption.
- Rough Time of Flight Estimation
- available bandwidth limits the precision of the time of flight estimates
- still vulnerable to On-the-fly and RF cycle slip attacks
- Random Phase Shifted Prover’s response
- prover introduces random (but agreed apriori with the verifier) phase shifts in its response
- forces the attacker to estimate the phase (can’t pre-calculate)
- does not prevent rollover attacks
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Possible Countermeasures
16
- Frequency Hopping
- nly valid if attacker is incapable of processing a wide band of frequencies
- simultaneously. Very weak assumption.
- Rough Time of Flight Estimation
- available bandwidth limits the precision of the time of flight estimates
- still vulnerable to On-the-fly and RF cycle slip attacks
- Random Phase Shifted Prover’s response
- prover introduces random (but agreed apriori with the verifier) phase shifts in its response
- forces the attacker to estimate the phase (can’t pre-calculate)
- does not prevent rollover attacks
Increases system complexity, makes phase ranging redundant
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Summary
17
- Investigated the vulnerability of carrier phase-based ranging system to
distance modification attacks
- Showed that it is possible to reduce the estimated distance to less than 3 m
even though the devices were more than 50 m apart
- Phase-based ranging systems are not suitable for security-critical
applications
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Summary
17
- Investigated the vulnerability of carrier phase-based ranging system to
distance modification attacks
- Showed that it is possible to reduce the estimated distance to less than 3 m
even though the devices were more than 50 m apart
- Phase-based ranging systems are not suitable for security-critical
applications
Are we really close? Verifying Proximity in Wireless Systems Aanjhan Ranganathan and Srdjan Capkun IEEE Security & Privacy Magazine, July-August edition, 2017
On the Security of Carrier Phase-based Ranging Aanjhan Ranganathan
Summary
17
- Investigated the vulnerability of carrier phase-based ranging system to
distance modification attacks
- Showed that it is possible to reduce the estimated distance to less than 3 m
even though the devices were more than 50 m apart
- Phase-based ranging systems are not suitable for security-critical
applications Thank you
www.aanjhan.com
Are we really close? Verifying Proximity in Wireless Systems Aanjhan Ranganathan and Srdjan Capkun IEEE Security & Privacy Magazine, July-August edition, 2017