We show that there exists a natural protocol problem which has a simple solution in the random-oracle (RO) model and which has no solution in the complexity-theoretic (CT) model, namely the problem of constructing a non-interactive communication protocol secure against adaptive adversaries a.k.a. non-interactive non-committing encryption. This separation between the models is due to the so-called programability of the random oracle. We show this by providing a formulation of the RO model in which the oracle is not programmable, and showing that in this model, there does not exist non-interactive non-committing encryption.
CITATION STYLE
Nielsen, J. B. (2002). Separating random oracle proofs from complexity theoretic proofs: The non-committing encryption case. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2442, pp. 111–126). Springer Verlag. https://doi.org/10.1007/3-540-45708-9_8
Mendeley helps you to discover research relevant for your work.