Re: New hash contest by NIST, similair to AES competition
- From: Thomas Pornin <pornin@xxxxxxxxx>
- Date: 26 Jan 2007 16:35:56 GMT
According to Alan <alan@xxxxxxxxxx>:
Symmetric block ciphers can be used in various ways to produce a
message digest:
Technically, Whirlpool is a block cipher (named "W"), derived from
Rijndael, used in such a way. And so are other hash functions. The block
cipher associated with SHA-1 is called SHACAL.
The troubles with this approach are:
-- The hash output size is equal to the block size; the standard AES has
a 128-bit block size, which is fine for a block cipher but too small,
by modern standards, for a hash function.
-- The data is fed to the block cipher as the key. Key schedule is often
quite slow in block cipher designs.
-- The characteristics needed for a proper hash function are unusual
for a block cipher. Namely, collision attacks for the hash function
are related key attacks on the block cipher. Related keys are seldom
studied, not much is known about them, and they are of little
significance to a block cipher when used for encryption. Related keys
were not a selection criterion for the processus which selected Rijndael
as being the AES.
Hence further research is needed.
--Thomas Pornin
.
- Follow-Ups:
- Re: New hash contest by NIST, similair to AES competition
- From: Mike Amling
- Re: New hash contest by NIST, similair to AES competition
- References:
- Prev by Date: Re: New hash contest by NIST, similair to AES competition
- Next by Date: Re: New hash contest by NIST, similair to AES competition
- Previous by thread: Re: New hash contest by NIST, similair to AES competition
- Next by thread: Re: New hash contest by NIST, similair to AES competition
- Index(es):
Relevant Pages
|
|