Privacy amplification secure against active adversaries
The notion of \textit{privacy amplification} is used to denote a technique where \textit{public discussion} (i.e. communication over an insecure channel) is used to transform a partially secret string into a highly secret one. Common scenario includes two parties \textit{Alice} and \textit{Bob} who share a string \(S\) and an opponent \textit{Eve} who has some partial knowledge about \(S\). By using appropriate protocol \textit{Alice} and \textit{Bob} communicate over an insecure channel and agree on another string such that \textit{Eve}'s knowledge about that string is negligible. Previously, privacy amplification has been analysed by \textit{C. H. Bennett, G. Brassard, C. Crépeau} and \textit{U. M. Maurer} [IEEE Trans. Inf. Theory 41, No. 6, Pt. 2, 1915-1923 (1995; Zbl 0856.94018)] under the ``passive adversary assumption, i.e. assuming that \textit{Eve} is not able to insert or modify messages without being detected by \textit{Alice} or \textit{Bob}. In the paper this restriction is removed, i.e. privacy amplification secure even against active adversary is considered. First, information-theoretically secure message authentication where the opponent is allowed to have some partial information about the shared key is investigated. It is shown that such authentication is possible under the sole assumption that no conditional probability of a certain key, given the opponent's information, exceeds precomputed bound. After that it is shown that privacy amplification secure against active adversary is possible and respective bounds for \textit{Eve}'s min-entropy about shared partially secret string and for the maximal length of the generated highly secure string are given. Finally, some open problems are briefly outlined. For the entire collection see [Zbl 0870.00047].
- Reusable fuzzy extractors for low-entropy distributions
- Multi-source non-malleable extractors and applications
- Privacy amplification from non-malleable codes
- Secrecy without perfect randomness: cryptography with (bounded) weak sources
- Amplifying Privacy in Privacy Amplification
- Privacy amplification with asymptotically optimal entropy loss
- Simultaneous secrecy and reliability amplification for a general channel model
- Privacy amplification for quantum key distribution
- An Improved Robust Fuzzy Extractor
- Key Agreement from Close Secrets over Unsecured Channels
- Strong security against active attacks in information-theoretic secret-key agreement
- Secret-key agreement over unauthenticated public channels-part III: privacy amplification
- Generalized privacy amplification
- Non-malleable extractors and non-malleable codes: partially optimal constructions
- Nonmalleable extractors and codes, with their many tampered extensions
- Privacy amplification with asymptotically optimal entropy loss
- Theory of Cryptography
- Deterministic extractors for small-space sources
- On secret sharing, randomness, and random-less reductions for secret sharing
- Extractors: low entropy requirements colliding with non-malleability
- Efficient randomized strong 2-source non-malleable extractor for any linear min-entropy
- Codes for adversaries: between worst-case and average-case jamming
- Error correcting codes, block designs, perfect secrecy and finite fields
This page was built for publication: Privacy amplification secure against active adversaries
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4364560)
