TY - GEN
T1 - Design of a True Random Number Generator Based on MRAM Devices
AU - Rios, Manuel Aguilar
AU - Jain, Saloni
AU - Cambou, Bertrand
N1 - Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2023
Y1 - 2023
N2 - Cryptographic schemes heavily rely on reliable random number generators. In a wide range of key generation schemes, such as the ones considered for post-quantum cryptography, cryptographic keys must be generated with random numbers that opponents cannot guess. Pseudo-random numbers (PRNs) are based on mathematical formulas, making them highly vulnerable to attacks. The generation of true random numbers (TRNs) can be achieved through various physical devices, including sensors, thermal noise generators, and many other components with high levels of stochasticity. A true random number generator (TRNG), even though it is slower, can generate non-deterministic data, making it highly secure. Magneto-resistive random access memories (MRAM), due to their high speed, non-volatility, high reliability, low fabrication cost, and low power consumption, are viable candidates for the design of cryptographic keys and true random numbers. The purpose of the work presented in this paper has been to exploit the physical characteristics of MRAMs to design true random numbers. We propose hardware and software implementations of TRNG schemes. We test random sequences using the statistical test suite from the National Institute of Standards and Technology (NIST) for random and pseudo-random numbers. Furthermore, we propose a post-processing method that uses XORing with pseudo-random numbers to improve randomness and security.
AB - Cryptographic schemes heavily rely on reliable random number generators. In a wide range of key generation schemes, such as the ones considered for post-quantum cryptography, cryptographic keys must be generated with random numbers that opponents cannot guess. Pseudo-random numbers (PRNs) are based on mathematical formulas, making them highly vulnerable to attacks. The generation of true random numbers (TRNs) can be achieved through various physical devices, including sensors, thermal noise generators, and many other components with high levels of stochasticity. A true random number generator (TRNG), even though it is slower, can generate non-deterministic data, making it highly secure. Magneto-resistive random access memories (MRAM), due to their high speed, non-volatility, high reliability, low fabrication cost, and low power consumption, are viable candidates for the design of cryptographic keys and true random numbers. The purpose of the work presented in this paper has been to exploit the physical characteristics of MRAMs to design true random numbers. We propose hardware and software implementations of TRNG schemes. We test random sequences using the statistical test suite from the National Institute of Standards and Technology (NIST) for random and pseudo-random numbers. Furthermore, we propose a post-processing method that uses XORing with pseudo-random numbers to improve randomness and security.
KW - Cryptographic Schemes
KW - Exclusive OR Logic (XOR)
KW - Low Power
KW - Magnetoresistive Random Access Memory (MRAM)
KW - Memory Array Components
KW - Non-Volatile
KW - Physical Unclonable Function (PUF)
KW - Pseudo-Random Number Generation
KW - Random Number Generation
KW - Ternary States
KW - True Random Number Generation
UR - http://www.scopus.com/inward/record.url?scp=85172254287&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85172254287&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-37963-5_90
DO - 10.1007/978-3-031-37963-5_90
M3 - Conference contribution
AN - SCOPUS:85172254287
SN - 9783031379628
T3 - Lecture Notes in Networks and Systems
SP - 1320
EP - 1341
BT - Intelligent Computing - Proceedings of the 2023 Computing Conference
A2 - Arai, Kohei
PB - Springer Science and Business Media Deutschland GmbH
T2 - Proceedings of the Computing Conference 2023
Y2 - 22 June 2023 through 23 June 2023
ER -