espanolEste trabajo propone un nuevo criptosistema, que combina el protocolo Diffie-Hellman en el cual se implementan curvas hiperelipticas sobre GF(2n), con la sincronizacion de Tree Paritty Machines (TPM). La seguridad propuesta para este criptosistema se centra en superar una debilidad de la sincronizacion neuronal. Especificamente, que el vector de estimulos es publico, lo cual permite a un atacante intentar sincronizar con uno de los participantes de la sincronizacion. Enfocandose en esta debilidad, existen los siguientes ataques: simple, genetico, geometrico y probabilistico. En el criptosistema propuesto, el vector de estimulo inicial se encuentra oculto, porque este vector se obtiene como la clave comun secreta en el protocolo Diffie-Hellman. Luego, en cada iteracion, los vectores de estimulo se mantendran en secreto. Esta condicion hace que el tiempo de aprendizaje tlear aumente en aproximadamente 115% con respecto al tiempo de sincronizacion tsync en promedio, cuando el criptosistema propuesto se compara con la sincronizacion de TPM clasica. EnglishThis work proposes a new cryptosystem, combining a Diffie-Hellman protocol in which hyperelliptic curves over GF(2n) are implemented, with a Tree Parity Machine (TPM) synchronization. Security proposed for this cryptosystem is focused on overcoming a weakness of neuronal synchronization. Specifically, the stimulus vector that is public, which allows an attacker to try to synchronize with one of the participants of the synchronization. Focusing on this weakness, there are the following attacks: genetic attack, geometric attack and probabilistic attack. In the proposed cryptosystem, the initial stimulus vector will be hidden, because this vector is obtained as the common secret key in the Diffie-Hellman protocol. Then in each iteration, the stimulus vectors will be kept secret. This condition causes the learning time tlear to increase by a term of approximately 115% regarding the synchronization time tsync on average when the proposed cryptosystem is compared to the classic TPM synchronization.
[1]
Andreas Ruttor,et al.
Dynamics of neural cryptography.
,
2007,
Physical review. E, Statistical, nonlinear, and soft matter physics.
[2]
T. S. B. Sudarshan,et al.
Neural Synchronization by Mutual Learning Using Genetic Approach for Secure Key Generation
,
2012,
SNDS.
[3]
Tanja Lange,et al.
Handbook of Elliptic and Hyperelliptic Curve Cryptography
,
2005
.
[4]
Wolfgang Kinzel,et al.
Cryptography based on neural networks—analytical results
,
2002
.
[5]
Wolfgang Kinzel,et al.
The Theory of Neural Networks and Cryptography
,
2003
.
[6]
Adi Shamir,et al.
Analysis of Neural Cryptography
,
2002,
ASIACRYPT.
[7]
Neal Koblitz,et al.
Algebraic aspects of cryptography
,
1998,
Algorithms and computation in mathematics.
[8]
Shu Lin,et al.
Error control coding : fundamentals and applications
,
1983
.
[9]
Christof Paar,et al.
Understanding Cryptography
,
2018,
Springer Berlin Heidelberg.
[10]
R. Tennant.
Algebra
,
1941,
Nature.
[11]
W. Kinzel,et al.
Secure exchange of information by synchronization of neural networks
,
2002
.
[12]
Simon Haykin,et al.
Neural Networks and Learning Machines
,
2010
.