PLUG-discuss digest, Vol 1 #2088 - 14 msgs

[Craig S.]

>
> Message: 10
> Date: Tue, 26 Mar 2002 14:07:41 -0700
> From: "John (EBo) David" <ebo@leml.la.asu.edu>
> Subject: Re: Fw: 1024-bit RSA keys in danger of compromise
> To: plug-discuss@lists.plug.phoenix.az.us
> Reply-To: plug-discuss@lists.plug.phoenix.az.us
>
> technomage wrote:
> >
> > I did some reading on that document and a little research into the hardware.
> > with current technology, the various government entities involved can build a
> > cracker that will do the jon in MINUTES for less than 2 nillion USD.
> >
> > time to start incorporating feistal arrays in new encryption algorythms.
>
> after rereading my earlier note I realized that I was not clear in
> meaning how much harder is it to crack 4K+ keys as opposed to 1k keys.
> If the complexity of the problem has been cracked then it is only a
> stop-gap.  Comming up with a new approach to the problem is what is
> really needed.
>
> feistal arrays? never heard of them....
>
>   EBo --
>
> -

this is still unclear. There are many factors that come into play when determining the strength of a key. For instance if I choose a key of 'abc' repeated 10 times then encrypted and a completely random (ok this is impossible in practice but we can get enough randomness to make deciphering the key difficult on todays technology) key with 30 unique characters the
unique char key will be much stronger than the repetitous key. Also what is being encrypted is a big factor in difficulty of compromise. Generally the more something repeats in a enciphering algorithm the easier it is to compromise. There are other things that come into play that I do not fully understand yet. If you are serious on learning security then I would
recommend starting with the book "Applied Cryptography". It isn't the end all be all text but it is a good start. Also there are some good newsgroups with very good FAQ's that discuss encryption. Also the practice of good encryption schemes relies on strength of the key vs. strength of the algorithm.

Feistel Arrays are more commonly known as Feistel Networks or GUFN's. It is a practice of block enciphering. That is a super simplified response and about all I know of them at this time.

Craig S.