Security adds an extra
design dimension
Ingrid Verbauwhede
KU Leuven COSIC
UCLA EE
Acknowledgement:
current and former
PhD students
Past: security for Embedded system
Old attack model (simplified view):
-Attack on channel between communicating parties
-Encryption and cryptographic operations in black boxes
-Protection by strong mathematic algorithms and protocols
Past: design for efficiency
Efficiency, flexibility, power, energy efficiency
Similar to other fields
DES processor
[EuroAsic 1991]
Rijndael/AES
[JSSC 2003]
Current: security for embedded system
Modified Attack Model (also simplified view):
-Attack channel and endpoints
-Encryption and cryptographic operations in gray boxes
-Protection by strong mathematic algorithms and protocols
-Protection by secure implementation
Need secure implementations not only algorithms
Design for efficiency AND security
SEMA attack: Simple Electromagnetic Attack on Elliptic Curve
Public Key implementation.
[E. Demulder EUROCON 2005]
Crypto engines with DPA countermeasures
Efficient, low power, side-channel resistance
Elliptic Curve within budget of passive RFID tag
Secure
Secure
[CHES2005]
Insecure
[DAC201?]
Future: security for immersed system:
which model?
New Model (also simplified view):
-Attack on the System
-Firewalls? IOT: there is NO inside versus outside
-Encryption, trust, security immersed
-Devices cooperate to build up trust
-All old requirements still stand
Challenge 1: Moores law helps (and hurts)
Moores law helps to implement more complex algorithms
with larger key (and word) lengths.
Moores law also hurts: gives more power to the attacker.
Demands on key lengths:
Challenge 2: IOT - Crypto diversity
[Terabits,
Long lasting
Security]
[Lightweight
HW entangled]
[Mbits
Security]
9
[Source photograph: J. Rabaey: A Brand New Wireless Day]
Challenge 3: Post-quantum crypto
Mathematical foundation of
existing public key algorithms
disappears
Quest for novel post-quantum
secure public key algorithms.
Lattice based: dimension
from N=256 to 2^15
Challenge 4: HW entangled security
Hardware Enabled CrypTO and Randomness
PUF
Secure Sketch:
Helper Data
Algorithm
(HDA)
1000s of
PUF bits
Universal /
Cryptographic
HASH
128bit key
11
Lightweight
Crypto
Algorithm
Research on True Random Numbers
12
Conclusions
Combine design for low power, low energy with design for
security
Design for security: as strong as weakest link!
IoT: what is inside, what is outside?
For the last 20 years: design for low power
[Chandrakasan 1992: Low Power CMOS design]
For the next 20 years: design for security