This article presents a new method of cryptographically resistant hashing information based on multivariate models. Since developed hash function is irreversible and resistant to collisions of the first and second kind, it could be used for cryptographic hashing of information in various data systems. Part of the proposed method hashing information is pseudorandom number generator based on multivariable models. The results show that the generator is close to the ideal. Accordingly, the computational complexity of finding collision close to 2n/2 and the hash function is cryptographically strong.
multivariate models, cryptographic hashing, cryptographic resistance, protection of information