A one-way function from thermodynamics and applications to cryptography (Q1879458)

The authors propose a new construction for a one-way function. The function is based on the second principle of thermodynamics and on the heat flow equation. The function might be used as a cryptographically strong primitive, and as the authors have suggested it might be applied in various ways; for example, to construct key exchange or password verification algorithms. Moreover, the function can be implemented on quantum computers. So when the first quantum computer is built and all classical cryptosystems based on computationally hard problems (like factorization, see RSA) are broken, the proposed function still can be used. The idea of cryptosystems based on the differential (heat) equation is very original and uncommon. It is a little similar to the idea of applying chaos theory to the construction of ciphers, which has appeared recently, but only for symmetric cases. Both of the above ideas are based on continuous functions and have the same problem in implementation on contemporary ``discrete'' computers. The authors point out the difficulties of the practical implementation related to this discretization problem. But it looks like these problem can be easily overcome. The idea of constructing cryptographic primitives based on thermodynamics looks quite promising. But we have to look at both sides of usefulness in cryptosystems: security and efficiency. So there is still the question about the performance of primitives constructed in this manner. All aspects of the performance need to be considered, like speed and size of implementation, secure sizes of the keys, etc. To replace currently used public key cryptosystems (like RSA or Diffie-Hellman), we need both more secure and faster algorithms. Whether these will rely on primitives based on thermodynamics remains an open problem. Much more analysis and investigation in this matter are required.