TY - GEN
T1 - Tamper Sensitive Ternary ReRAM-Based PUFs
AU - Cambou, Bertrand
AU - Chen, Ying Chen
N1 - Publisher Copyright:
© 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2021
Y1 - 2021
N2 - The physical properties of resistive random-access memories can be exploited to design physical unclonable functions, they can also be used to insert sensing elements detecting tampering activities. We are presenting how un-filtered cryptographic key generation cycles can permanently damage about 20% to 40% of the memory cell population, while the strong cells return to their pristine state after operation. During an enrollment cycle performed once upfront, the weak cells are identified, tracked with a ternary state, and never used during subsequent key generation cycles thereafter. The weak cells are likely to be damaged when the opponent blindly characterizes the physical unclonable functions or attempt to generate cryptographic keys. Thereby these weaker cells act as sensing elements of this type of attacks. The cryptographic scheme is optimized to tolerate a certain level of failure of the stronger cells, and to compute ternary states to keep track of the weaker cells. The implementation was written in Phyton and C++ at the server level, and in C at the client level. The protocols use the standard hash algorithm SHA-3, and the extended output function SHAKE.
AB - The physical properties of resistive random-access memories can be exploited to design physical unclonable functions, they can also be used to insert sensing elements detecting tampering activities. We are presenting how un-filtered cryptographic key generation cycles can permanently damage about 20% to 40% of the memory cell population, while the strong cells return to their pristine state after operation. During an enrollment cycle performed once upfront, the weak cells are identified, tracked with a ternary state, and never used during subsequent key generation cycles thereafter. The weak cells are likely to be damaged when the opponent blindly characterizes the physical unclonable functions or attempt to generate cryptographic keys. Thereby these weaker cells act as sensing elements of this type of attacks. The cryptographic scheme is optimized to tolerate a certain level of failure of the stronger cells, and to compute ternary states to keep track of the weaker cells. The implementation was written in Phyton and C++ at the server level, and in C at the client level. The protocols use the standard hash algorithm SHA-3, and the extended output function SHAKE.
KW - Cyber-physical systems
KW - Intrusion detection
KW - Secure protocols
UR - http://www.scopus.com/inward/record.url?scp=85112734600&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85112734600&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-80129-8_67
DO - 10.1007/978-3-030-80129-8_67
M3 - Conference contribution
AN - SCOPUS:85112734600
SN - 9783030801281
T3 - Lecture Notes in Networks and Systems
SP - 1020
EP - 1040
BT - Intelligent Computing - Proceedings of the 2021 Computing Conference
A2 - Arai, Kohei
PB - Springer Science and Business Media Deutschland GmbH
T2 - Computing Conference, 2021
Y2 - 15 July 2021 through 16 July 2021
ER -