On Nov 17, 5:34 pm, smu johnson <[email protected]> wrote:
> Hi,
>
> Sorry I am not much of a C++ expert so instead of pulling my hair out I'd
> like to ask this question:
>
> Does it take a long time to do DH bignum math (powMod) with sizes higher
> than 8192 bit prime groups?  The RFC spec says it will take an impractical
> amount of time @http://tools.ietf.org/html/rfc3526, but I am thinking
> maybe the times have changed with faster computers.
I believe Crypto++ uses Montgomery, which I think is k(k+1) or k^2.
You'll spend most of the time looking for the safe prime.

> The reason is I want to securely do DH for 256 bit symmetric keys, and the
> RFC seems to think that you'll need very high prime groups to do it.
You will need a modulus around 15360 (NIST). ECRYPT II recommends
similar. See http://www.cryptopp.com/wiki/Security_level and ECRYPT II
Yearly Report on Algorithms and Keysizes.

> Also, is there a limit for the group size in Crypto++?
The limit is based on the size of class Integer, which is something
like 2^256^256. In practice, there is no limit.

> Does it have the RFC groups hardcoded in them, or must you generate
> your own primes for its DH?
I'll get some example up soon. Check 
http://www.cryptopp.com/wiki/Diffie-Hellman.

Jeff

-- 
You received this message because you are subscribed to the "Crypto++ Users" 
Google Group.
To unsubscribe, send an email to [email protected].
More information about Crypto++ and this group is available at 
http://www.cryptopp.com.

Reply via email to