Limits on the Efficiency of One-Way Permutation-Based Hash Functions

  • Authors:
  • Jeong Han Kim;Daniel R. Simon;Prasad Tetali

  • Affiliations:
  • -;-;-

  • Venue:
  • FOCS '99 Proceedings of the 40th Annual Symposium on Foundations of Computer Science
  • Year:
  • 1999

Quantified Score

Hi-index 0.00

Visualization

Abstract

Naor and Yung have shown that a one-bit-compressing universal one-way hash function (UOWHF) can be constructed based on a one-way permutation. This construction can be iterated to build a UOWHF which compresses by \math bits, at the cost of \math invocations of the one-way permutation. We show that this construction is not far from optimal, in the following sense: there exists an oracle relative to which there exists a one-way permutation with inversion probability \math (for any \math, but any construction of an \math-bit-compressing UOWHF requires \math invocations of the one-way permutation, on average. (For example, there exists in this relativized world a one-way permutation with inversion probability \math, but no UOWHF that invokes it fewer than \math times.) Thus any proof that a more efficient UOWHF can be derived from a one-way permutation is necessarily non-relativizing; in particular, no provable construction of a more efficient UOWHF can exist based solely on a "black box" one-way permutation. This result can be viewed as a partial justification for the practice of building efficient UOWHFs from stronger primitives (such as collision-intractable hash functions), rather than from weaker primitives such as one-way permutations.