Re: Secret sharing algorithm with chosen keys
- From: usenet@xxxxxxxxxxxxxxxxxxxxxxxxxxx (James Taylor)
- Date: Wed, 25 Feb 2009 06:00:59 +0700
Lethal Possum <lethal.possum@xxxxxxxxx> wrote:
James, concerning your first suggestion: I am not sure it would be
secure. When you say "instead of public keys, you'd have the XORed
token set for each T people", now let's say we have N = 3 and T = 2.
Let's call the keys A, B and C. Let's call the symmetric key S. If I
understood your proposition correctly, you compute A + B + S, A + C +
S and B + C + S and put that with the message encrypted by S.
Yes, that's the idea.
The issue is that for example (A + B + S) + (A + C + S) + (B + C + S)
= S. In other words you can derive the main symmetric encryption key
from the decryption keys.
Hmmm, I see the problem.
But I'll have to think is there is a way to make it work
maybe using something less reversible than XOR to create
the decryption keys.
Perhaps each "decryption key" as you call it, should be a ciphered copy
of the symmetric key S and this cipher should be keyed on the XORed
combination of the T tokens. That way, even knowing all the decryption
keys doesn't help you derive the symmetric key without actually having
the relevant XORed combination to reverse the cipher protecting a
decryption key.
Hope that helps. Unfortunately, the other responses went over my head,
but I feel confident they are worth considering in more depth than my
novice ideas.
--
James Taylor
.
- References:
- Secret sharing algorithm with chosen keys
- From: lethal . possum
- Re: Secret sharing algorithm with chosen keys
- From: Scott Fluhrer
- Re: Secret sharing algorithm with chosen keys
- From: James Taylor
- Re: Secret sharing algorithm with chosen keys
- From: Scott Fluhrer
- Re: Secret sharing algorithm with chosen keys
- From: Lethal Possum
- Secret sharing algorithm with chosen keys
- Prev by Date: Re: Paper & pencil password algorithm
- Next by Date: Re: Security Review Summary of NIST SHA-3 Round 1
- Previous by thread: Re: Secret sharing algorithm with chosen keys
- Next by thread: Re: Secret sharing algorithm with chosen keys
- Index(es):
Relevant Pages
|