Re: New hash contest by NIST, similair to AES competition

"Alan" <alan@xxxxxxxxxx> wrote
news:1169827591.056794.255390@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Symmetric block ciphers can be used in various ways to produce a
message digest:
http://en.wikipedia.org/wiki/Hash_functions_based_on_block_ciphers
Is there a drawback to that approach (other than wanting more diversity
in the toolbox)?


Symmetric block ciphers or block cipher technology is very good starting
point, but the classical block ciphers are not. In the introduction of "A
New Concept of Hash Functions SNMAC Using a Special Block Cipher and
NMAC/HMAC Constructions", IACR ePrint archive Report 2006/376 , October,
2006, we wrote:

Block ciphers should be used in hash functions in another way than we have
seen so far. We call them special block ciphers (SBC) and we formulate their
properties. This new cryptographic primitive surpasses the classical
conception of a block cipher. The basic property of SBC is that an attacker
can fully control its key. The block ciphers have not been designed with
this requirement, yet. Therefore, contemporary block ciphers are not too
suitable for being used in hash functions. We have to subordinate the design
of these block ciphers to the aforesaid new demand that the attacker has
full control over the plaintext and the key.
Vlastimil


.

Relevant Pages

  • Re: What exactly is avalanche effect?
    ... In cryptography, the avalanche effect refers to a desirable property ... typically block ciphers and cryptographic ... `and cryptographic hash functions'. ... As far as I can see, `avalanche effect' refers to the idea that changing ...
    (sci.crypt)
  • Re: Collision in SHA-0 - concerns for SHA-1
    ... >>on e.g. block ciphers in some clear way rather than simply ... >>sufficient accompanying documentation of designs. ... hash functions using symmetric block algorithms'. ... rather blind trust on authorities and that's not very good. ...
    (sci.crypt)
  • Re: Variation on CTR mode
    ... >> based on block ciphers? ... > And what about keyed hash functions obtained encrypting a CRC ...
    (sci.crypt)
  • Re: Variation on CTR mode
    ... > Do you have similar opinions on hash functions that are ... > based on block ciphers? ... And what about keyed hash functions obtained encrypting a CRC ...
    (sci.crypt)
  • Re: use of randomized components in encryption algorithms
    ... component when encrypting data? ... These are block ciphers, and as such, are completely deterministic. ... To create cryptosystems from these block ciphers, ... If alternatives do the better things design wise, ...
    (sci.crypt)
Quantcast